52 pages including cover

Knowledge Digest for IT Community

50/-
0970-647X

Volume No. 41 | Issue No. 10

7 | |October
January2017
2018
0970-647X
ISSN ISSN

machine
intelligence
www.csi-india.org
www.csi-india.org

Cover Story Article

Machine Learning 7 Make friendship with your Machine :
Cover Story RESEARCH FRONT
The HUMachine 26
CSI Nihilent eGovernance Awards 7 Remote Monitoring and Localization
TECHNICAL TRENDS using Tools for e-Governance 17
Sensors: CORNER
SECURITY
A Roadmap for Biological Network Analysis 9 Cyber Crime and Social Security –
TECHNICAL TRENDS
RESEARCH FRONT The Dark side of Internet 37
Article
Meri Sadak 2.0Development
Collaborative : of Electro-Optical Ontology Modeling in E-Governance for a
One stepEvaluation
Payload closer to SMART 15 xpress) 14
CITY(XSCoPE
Software Semantic Digital India 25
Know Your CSI
Executive Committee (2017-18/19)
President Vice President Hon. Secretary
Mr. Sanjay Mohapatra Mr. Gautam Mahapatra Prof. A. K. Nayak
D/204, Kanan Tower, Vailla No: 8, Maithri Enclave, Director, Indian Institute of Business
Patia Square, Bhubaneswar Near Tulsi Gardens, Yapral Kapra, Management, Budh Marg, Patna
Email : [email protected] Hyderabad-500 062. (E) [email protected]
(E) [email protected]

Hon. Treasurer Immd. Past President

Mr. Manas Ranjan Pattnaik Dr. Anirban Basu
Plot No. N-24, 25 Chandaka Indl. 309, Ansal Forte, 16/2A,
Estate, Patia, KIIT, Bhubaneswar Rupena Agrahara, Bangalore
(E) [email protected] Email : [email protected]

Regional Vice-Presidents
Region-I Region-II Region-III
Mr. Arvind Sharma Mr. Devaprasanna Sinha Prof. Vipin Tyagi
3/294, Vishwas Khand, 73B Ekdalia Road, Dept. of CSE, Jaypee University of
Gomati Nagar, Lucknow-226010. UP Kolkata - 700 019 Engg. and Tech., Guna - MP
(E) [email protected] (E) [email protected] (E) [email protected]

Region-IV Region-V Region-VI

Mr. Hari Shankar Mishra Mr. Vishwas Bondade Dr. Shirish S. Sane
Command Care, Opp. Loreto Convent No. 774, 2nd Stage, Indiranagar, Vice-Principal, K K Wagh Institute of
School, A. G. Office Road, Doranda, Bangalore 560038 Engg Education & Research,Nashik,
Ranchi – 834002, Jharkhand (E) [email protected] (E) [email protected]
(E) [email protected]

Region-VII
Dr. M. Sundaresan
Professor and Head,
Dept. of IT, Bharathiar University,
Coimbatore - 641046, Tamil Nadu.
(E) [email protected]

Division Chairpersons
Division-I Division-II Division-III
Mr. Apoorva Agha Prof. P. Kalyanaraman Mr. Raju L. kanchibhotla
8, Katra Road, Allahabad, UP - 211002 Plot No. 139, Vaibhav Nagar, Aashirvad, 42/260/1/2,
(E) [email protected] Phase I, Opp VIT Gate 3, Vellore – 632014. Shramik Nagar,
(E) [email protected] Moulali Hyderabad-500 046, India
(E) [email protected]
Division-IV Division-V
Dr. Durgesh Kumar Mishra Dr. P. Kumar
H-123-B, Vigyan Nagar, Professor and Head
Dept. of Computer Science and Engineering,
Annapurna Road, Indore Rajalakshmi Engineering College, Chennai – 602 105.
(E) [email protected] (E) [email protected]

Nomination Committee (2017-2018)

Prof. K. Subramanian Dr. Brojo Kishore Mishra Mr. Subimal Kundu
B 28,Tarang Apmts, Associate Professor, Dept. of IT, Flat No. 1A, Block - 7,
Plot 19, IP Extn, C. V. Raman College Engineering, Space Town Housing Complex,
Bhubaneshwar - 752054. India P.O. Airport, Kolkata – 700052
Patparganj, Delhi - 110092 (E) [email protected] (E) [email protected]
(E) [email protected] [email protected] [email protected]

an individual. CSI Headquarter :

Samruddhi Venture Park, Unit No. 3,
2 are friends. 4th Floor, MIDC, Andheri (E),
Mumbai-400093, Maharashtra, India
3 is company. Phone : 91-22-29261700
Fax : 91-22-28302133
more than 3 makes a society. The Email : [email protected]
arrangement of these elements makes
the letter ‘C’ connoting ‘Computer CSI Education Directorate : CSI Registered Office :
Society of India’. CIT Campus, 4th Cross Road, Taramani, 302, Archana Arcade, 10-3-190,
Chennai-600 113, Tamilnadu, India St. Johns Road,
the space inside the letter ‘C’ connotes Phone : 91-44-2254 1102-03 Secunderabad-500025,
an arrow - the feeding-in of information Fax : 91-44-2254 2874 Telengana, India
or receiving information from a computer. Email : [email protected] Phone : 91-40-27821998
 CSI  Communications Volume No. 41 • Issue No. 10 • january 2018

Contents
Chief Editor
S S Agrawal
KIIT Group, Gurgaon

Editor Cover Story

Prashant R Nair Machine Learning 7
Amrita Vishwa Vidyapeetham, Coimbatore Chandu Vaidya and Shubham Uttarwar
Published by Technical Trends
AKSHAYA KUMAR NAYAK A Roadmap for Biological Network Analysis 9
Hony. Secretary Nikhila T Suresh, Sreeja Ashok and U. Krishnakumar
For Computer Society of India Blockchain : A Distributed Ledger Technology 12
V. Sai Chandrahaas and V. Ravi Sankar
Editorial Board:
Arun B Samaddar, NIT, Sikkim Research Front
Collaborative Development of Electro-Optical Payload Evaluation Software (XSCoPExpress) 14
Bhabani Shankar Prasad Mishra,
KIIT University, Bhubanewar Nutan Kumari, Jitendra Sharma and Amit Dave
Debajyoti Mukhopadhyay, MIT, Pune
Text Detection and Recognition in Natural Scene Images 19
Sharath Kumar Y H and Puneeth P
J Yogapriya, Kongunadu Engg. College, Trichy
M Sasikumar, CDAC, Mumbai, Articles
Make friendship with your Machine : The HUMachine 26
R Subburaj, SRM University, Chennai Anila M.
R K Samanta, Siliguri Inst. of Tech., West Bengal Imaging Applications using Machine Intelligence 28
R N Behera, NIC, Bhubaneswar Akila K. and Jayasri P.
Sudhakar A M, University of Mysore
Understanding Machine Vision by Human Activity Recognition 31
Dushyant Kumar Singh
Sunil Pandey, ITS, Ghaziabad Role of Machine Intelligence for sustainable agricultural development 33
Shailesh K Srivastava, NIC, Patna Mahendra Balkrishna Salunke and Ashish V. Salunkhe
Vishal Mehrotra, TCS Machine Intelligence Challenges in Military Robotic Control 35
S Balakrishnan and D Deva

Security Corner
Design, Print and Dispatch by Cyber Crime and Social Security – The Dark side of Internet 37
Priyanka Tomar
GP Offset Pvt. Ltd.
PLUS
Know Your CSI 2nd Cover
Computer Society of India & Patna Management Association jointly organize 6
One-day seminar on Cashless Economy
Please note: Prof. Mike Hinchey, President, IFIP has agreed to grace CSI Annual Convention 6
CSI Communications is published by Computer
Society of India, a non-profit organization. at Kolkata as Keynote Speaker
Views and opinions expressed in the CSI
Communications are those of individual
Prospective Contributors of CSI Communications 11
authors, contributors and advertisers and they IEEE MGA Leadership Award to Mr. H R Mohan 18
may differ from policies and official statements
of CSI. These should not be construed as legal CSI Nihilent eGovernance Awards 2017 – Invitation 25
or professional advice. The CSI, the publisher,
the editors and the contributors are not
National Conference on Future Innovations in 41
responsible for any decisions taken by readers Computing Technologies and Machine Learning (FICTML-17)
on the basis of these views and opinions.
Although every care is being taken to ensure
CSI Mumbai Chapter – CSI-IT 2020 – A Report 42
genuineness of the writings in this publication, Team CSI 2017 in the ExecCom and Review Meeting held at Kolkata 43
CSI Communications does not attest to the
originality of the respective authors’ content. CSI Reports 45
© 2012 CSI. All rights reserved. Student Branch News 47
Instructors are permitted to photocopy isolated
articles for non-commercial classroom use CSI Calendar 2018 3rd Cover
without fee. For any other copying, reprint or
republication, permission must be obtained
CSI Invites Project Proposals from Faculty Members and Students Back Page
in writing from the Society. Copying for other Printed and Published by Akshaya Kumar Nayak on behalf of Computer Society of India, Printed GP Offset Pvt. Ltd.
than personal use or internal reference, or of
articles or columns not owned by the Society
269, 2nd Floor, A-2, Shah & Nahar Indl. Estate, Sitaram Jadhav Marg, Lower Parel, Mumbai 400 013 and Published from
without explicit permission of the Society or Samruddhi Venture Park, Unit No. 3, 4th Floor, Marol Industrial Area, Andheri (East), Mumbai 400 093. • Email : [email protected]
the copyright owner is strictly prohibited. Editor: Prashant R. Nair

3
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
Editorial
Dear Fellow CSI Members,
At the outset, we wish all our esteemed CSI members, a happy, prosperous and productive New Year 2018.
The theme for the Computer Society of India (CSI) Communications (The Knowledge Digest for IT Community)
January 2018 issue is Machine Intelligence.
“Artificial intelligence will reach human levels by around 2029. Follow that out further to, say, 2045, we will have
multiplied the intelligence, the human biological machine intelligence of our civilization a billion-fold.”
Prof. (Dr.) S. S. Agrawal
Chief Editor
Ray Kurzweil

After a series of thematic issues focusing on ICT in applications such as education, governance, agriculture and
health, CSI Communications is focusing on machine intelligence in this issue after issues on the research topics
of cyber physical systems and machine learning.
Machine intelligence is intelligence displayed by machines. In other words, any device that perceives its
environment and takes actions that maximize its chance of success at some goal. Machines are becoming
increasingly capable and performing high-end tasks such as understanding human speech, competing at a
high level in strategic game systems, autonomous cars, intelligent routing in content delivery networks, military
Prof. Prashant R. Nair simulations, and interpreting complex data, including images and videos.
Editor
The Cover story in this issue is Machine Learning by Chandu Vaidya & Shubham Uttarwara. In the cover story,
the authors have traced the evolution, tools and applications of machine intelligence.
The technical trends showcased are “A Roadmap for Biological Network Analysis” by Nikhila T Suresh, Sreeja
Ashok and U. Krishnakumar and “Blockchain : A Distributed Ledger Technology” by V. Sai Chandrahaas & V.
Ravi Sankar
In Research front, we have 2 contributions, “Collaborative Development of Electro-Optical Payload Evaluation
Software (XSCoPExpress)” by Nutan Kumari, Jitendra Sharma and Amit Dave from Indian Space Research
Organization (ISRO) and “Text Detection and Recognition in Natural Scene Images” by Sharath Kumar Y.H. and
Puneeth P.
Other articles in this issue on Machine Intelligence provide us information on its applications in image processing,
video processing, agricultural development and military robotic control. The articles are:
¬¬ “Make friendship with your Machine :The HUMachine” by Anila M.
¬¬ “Imaging Applications using Machine Intelligence” by Akila K. and Jayasri P.
¬¬ “Understanding Machine Vision by Human Activity Recognition” by Dushyant Kumar Singh
¬¬ “Role of Machine Intelligence for sustainable agricultural development” by Mahendra Balkrishna Salunke
and Ashish V. Salunkhe
¬¬ “Machine Intelligence Challenges in Military Robotic Control” by S Balakrishnan and D Deva

The Security Corner has 1 contribution, Cyber Crime and Social Security – The Dark side of Internet by Priyanka
Tomar

This issue also contains reports of various premier CSI conferences such as CSI IT 2020, CSI activity reports from
chapters & student branches, call for R&D funding proposals and calendar of events

We are thankful to entire ExecCom for their continuous support in bringing this issue successfully.

We wish to express our sincere gratitude to the CSI publications committee, editorial board, authors and
reviewers for their contributions and support to this issue.

We look forward to receive constructive feedback and suggestions from our esteemed members and readers at
[email protected].

With kind regards,

Prof. (Dr.) S. S. Agrawal, Chief Editor Prof. Prashant R. Nair, Editor

www.csi-india.org
4
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 Computer Society of India
TM

http://www.csi-india.org

Mr. Sanjay Mohapatra Shri.Gautam Mahapatra Prof. A. K. Nayak Mr. Manas Ranjan Pattnaik Dr Anirban Basu
President Vice President Hon. Secretary Hon. Treasurer Immd. Past President

President’s Message
From : President, Computer Society of India
Date : 01 January, 2018
Email : [email protected] / Cell : (91) 9861010656

Wish all our esteemed CSI members and their families a Happy present situation of India and
and Prosperous New Year 2018! Africa towards ICT Growth,
Opportunities and ways forward
International Summit: 2017 India-Africa ICT Summit @ NCR
and stressed upon need of
Region
developing India and Africa
bilateral relations through Information and Communication
Technologies. A suggestion was given to have a delegation
meet visit to Uganda for analyzing the grass root level issue
and Challenges and ways forward with the support of Bilateral
resource sharing during Mid February 2018. This summit was
an excellent opportunity for CSI to expand its international
collaborative footprint. CSI President also addressed all
delegates on the theme of “Setting up of IT Parks for promoting
IT-ITeS business in Africa”.
CSI 2017 Convention @ Kolkata
CSI ExecCom met at Kolkata on 9 December 2017 to review the
preparations for the CSI 2017 annual convention scheduled at
Science City, Kolkata from 19 to 21 January, 2018. The theme
of the convention is “Social Transformation- Digital Way” and
full-fledged preparations are on by the CSI 2017 organizing
team and Kolkata chapter for making it a grand success. After
rigorous review and quality check, 64 papers have been short-
listed for presentation and the proceedings will be published
by our partner, Springer Nature. The organizing team has put
together an impressive array of distinguished speakers from
top universities, institutions and companies such as IBM, Tata
Consultancy Services (TCS), Microsoft, Kansas State University
USA, Nanyang Technological University (NTU) Singapore, IISc
Bangalore, IIT Kharagpur, ISI Kolkata, Jadavapur University,
Saha Institute of Nuclear Physics, Internet Corporation
for Assigned Names and Numbers (ICANN), CDAC etc. It is
my humble request that all our CSI members from industry,
student branches etc register and participate in large numbers
and make this another historic convention for CSI.
Membership growth & CSI portal
Sustainable growth of membership is going on and
Mr. Sanjay Mohapatra, President, CSI represented CSI at the membership data validation process online is in progress.
2017 India Africa ICT Summit & Excellence Award conducted at CSI portal is presented with a new look and all dynamic &
Hotel Ramada, Gurgaon Central on 1 & 2 December, 2017. CSI status pages are designed by CSI Vice-President, Dr. Gautam
partnered with Innovation society India, Ministry of Electronics Mohapatra & Team. Dr. Mohapatra, despite his very extremely
and Information Technologies, WAIMS and Public Opinion busy schedule at DRDO is devoting a good quantum of time for
International Uganda in organizing the summit with the theme, CSI portal.
“Confronting ICT and Cyber Challenges, Developing India Please write your valuable ideas for growth of CSI at president@
and Africa through ICT, Opportunities, Challenges and way csi-india.org
forward”. CSI President participated in the inaugural ceremony
along with Mr. Saurabh Tiwari, Dy. Director, Department of With kind regards
Telecommunication, Govt. of India; Mr. N. K. Goyal, President
CMAI; African Delegation headed by Justice Ralph Ochan,
Chairman, Public Service Commission Uganda and others. Sanjay Mohapatra
In the inaugural ceremony, CSI President emphasized upon President, CSI
5
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 Computer Society of India & Patna
Management Association jointly organize
One-day seminar on Cashless Economy
Patna Management Association
with technical collaboration of
Computer Society of India organised
a one-day seminar on ‘Cashless
Economy for New India’ on 25th
November, 2017 at Dr. Zakir Hussain
Institute Auditorium, Patna. The
Seminar was inaugurated by Hon’ble
Mr. A N Singh, Former speaker of Bihar
Legislative Council. He said that, the
drive of digitalization in economy has
led to inclusive growth and reduction
in corruption. The Guest of honor Dr.
Dolly Sinha, Pro Vice Chancellor of
Patna University told that the drive of
digital money transfer has brought
huge amount of black money back into
the system.
Presenting the Key note address
in the Seminar, Prof. A K Nayak, the
Director of Indian Institute of Business
Management & National Secretary of the present govt. for setting up a Management Association Proposed
Computer Society of India threw light proper mechanism to implement the Vote of thanks. Others Who were
on various technical aspects of the cashless economy. Mr. Arun Kumar, Present, prominent among them
ecosystem of digital economy. He also former General Manager of Bank of Dr. Bhawana sinha, Vice Chairman
highlighted about exponential growth Baroda, talked about various practical CSI Patna Chapter & Head MCA
in cashless transaction after the de aspects and suggested few points for Dept. Patna Women’s College, Prof.
monetization process & explained successful implementation of this Ganesh Panday, Dy. Director, IIBM
about growth in e-transaction in drive. Patna & Former Bihar State Student
various sector In the beginning, Mr. S K N Haider, Coordinator, CSI & Mr. Manoj Sinha,
In his presidential address, Prof. President of Patna Management Vice President, Patna Management
U K Singh, Director General of IIBM Association welcomed the guests.Prof Association.
&ZHI group of institutions praised Alok Kumar, Hony Secretary, Patna n

Prof. Mike Hinchey, President, IFIP has agreed to grace

CSI Annual Convention at Kolkata as Keynote Speaker
Mike Hinchey is President of IFIP NASA “Expert”.
(International Federation for Information Hinchey holds a BSc in Computer Systems from
Processing), founded by UNESCO University of Limerick, an MSc in Computation from University
in 1960, and Professor of Software of Oxford and PhD in Computer Science from University of
Engineering at University of Limerick, Ireland. He is Chair Cambridge. He is a Fellow of IET, British Computer Society,
of IEEE UK and Ireland for 2018-2019. Irish Computer Society (of which he is also Vice President),
Until end of 2016, for 8 years he was Director of Lero-the Engineers Ireland and Engineers Australia. He is the author
Irish Software Research Centre, a multi-university research or editor of more than 20 books, including the forthcoming
centre with a footprint in each of Ireland’s universities. “Software Technology: 10 years of innovation in IEEE
Previous positions include Director of the Software Computer” (Wiley, 2018), more than 200 papers, and holds
Verification Research Centre at University of Queensland, 26 patents. He is Editor-in-Chief of Innovations in Systems
and Director of the NASA Software Engineering Laboratory and Software Engineering: a NASA Journal” (Springer) and
at Goddard Space Flight Center; he has also served as a Journal of the Brazilian Computer Society (Springer).

www.csi-india.org
6
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 COVER STORY

Machine Learning
Chandu Vaidya Shubham Uttarwar
Assistant professor, Rajiv Gandhi College of Engineering & Research Rajiv Gandhi College of Engineering & Research

1. Machine Learning – An can account for each one of them ahead we want to make predictions based
Introduction of time. Take for example, what if a on a set of examples. Unsupervised
[1] Machine learning can be broadly certain leaf is obstructed or flower is of machine learning problems are
defined as computational methods a different color or the shape is totally problems where our data does not
using the experience to improve the different from what we expected. We have a set of defined set of categories,
performance or to make accurate just simply cannot code it beforehand but instead, we are looking for the
predictions. In layman’s term, we and thus we have to use machine machine learning algorithms to help us
define machine learning as a series learning to solve this problem. To solve organize the data. Put in another way –
of mathematical manipulations any machine learning problem there supervised machine learning problems
performed on important data in order to are 4 steps that need to be followed. have a set of historical data points
gain valuable insights. More specifically, which we want to use to predict the
We Collect Data for one. Secondly,
it is the study of algorithms that learn future, unsupervised machine learning
we pick the model. Third, we train
from examples and experience instead problems have a set of data which we
the model and then finally we test the
of hardcoded rules. ML is evolved are looking for machine learning to
model. By simply adding the data to the
from the study of pattern recognition help us organize or understand.Again,
model it will start recognizing patterns
and computational learning theory in there is a further classification of both
for us. Now the data sets can come
artificial intelligence, machine learning supervised and unsupervised problems
in the whole bunch of different forms
explores the study and construction which can be studied. Reinforcement
namely csv, pdf, holograms, images but
of algorithms that can learn from includes a specific task or goal that the
should be in a well-known format which
and make predictions on data – such system must complete. Throughout the
we can parse from. The dataset then
algorithms overcome following strictly process, it receives feedback in order
has different features according to our
static program instructions by making to learn the desired behaviors. For
requirement. The sign of good features
data-driven predictions or decisions by example, the system encounters an
is that they should be simple for one.
building a model from sample inputs. error while performing the action or a
Secondly, they should be independent
Machine learning is closely and third they should be informative. reward for achieving the most favorable
related to and often overlaps with Then we choose one of the different outcome. Thus, the program is able to
computational statistics, which also models that are available already for learn the most effective approach via
focuses on prediction-making through machine learning. Choosing a model “reinforcement signals”.
the use of computers. It has strong ties can be tiresome, however, we can 2. Tools available:
to mathematical optimization, which narrow our search by determining the To use machine learning in your
delivers methods, theory and application type of problem we are solving and how application or even to learn it, there are
domains to the field. Ml involves large our dataset is and what are our two ways. First being, learning how to
learning from different experiences other requirements on the whole. Again, use libraries that act as black box i.e
and then using those experiences to its a matter of choice. Since we are they provide each functionality, but you
predict the correct outcomes in the classifying whether an image is of iris just cannot know what is happening
later stages of its use. [1]To define it flower or not, ours is the classification inside. Secondly, to learn how to write
in a computational language we say problem and thus we use a classifier. algorithms and find coefficients, fit the
that, “A computer program is said to Now that we have our dataset and model, find optimization points and
learn from experience E with respect to model, we train the model. Training much more so that you can curate your
some class of tasks T and performance is the step where the actual learning application as per your requirement.
measure P, if its performance at tasks occurs as our model iterates through The latter requires a pre-requisite of a
in T, as measured by P, improves with the dataset, it gets better and better at stronghold on calculus and probabilistic
experience E”. prediction over time. And we are done! theory and many such mathematical
For example, if we need to create an Next time, when we provide our model concepts. Though it might look a
application that recognizes a specific image of any flower it will be able to formidable task, but it’s the best way
species of the flower named iris. In predict whether the flower supplied is to learn machine learning, because
absence of machine learning, we iris or not. then you get to understand how things
would have to write a bunch of different [3] Generally, there are three main work behind the scene. However, if
functions to detect all the different types of machine learning problems: you just want to play along, there are
features of the iris flower. However, the supervised, unsupervised and a number of libraries and application
problem is that there is a large number reinforcement. Supervised machine programming interfaces that can get
of corner cases and there is no way we learning problems are problems where you your job done. To name a few good
7
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 COVER STORY

libraries out there, there is TensorFlow and public safety have a specific our future. It affects how we work,
by Google, Scikit-Learn (it comes need FOR Ml, as they have multiple socialize and play. For example, after
preloaded with a ton of datasets, which data sources, which can be mined one week of watching your favorite
you can use to play around a get a feel for identifying useful patterns and movies on Netflix, but when you sit
of the whole process). insights. For example sensor, data down next time to watch a movie on
3. Areas of Application can be analyzed to identify ways to Netflix, you realize that NetFlix actually
Giants like Google, Microsoft, Uber, minimize costs and increase efficiency. recommends movies that you will enjoy
Amazon, Facebook, Instagram, Apple Furthermore, ML can also be used watching. Machine learning here is
constantly use machine learning to gain to minimize identity thefts and detect helping create this experience for you
valuable insights from data, recognize fraud. by recommending movies you might
new trends and thus use them to 4. Healthcare love. Similar is the case on Amazon,
increase their customer base as well as With the advent of wearable next time you buy something, Amazon
profits. [2]The value of machine learning sensors and devices that use data automatically recommends you things
technology has been recognized by to access the health of a patient in that it thinks you might be interested in,
companies not only in technology real time, ML is becoming a fast- based on your past purchases.
sector but across several industries growing trend in healthcare. Sensors Again, when you are out
that deal with huge volumes of data. By in wearable provide real-time patient somewhere, having a gala night with
leveraging insights obtained from this information, such as overall health your friends and it’s late in the night,
data, companies are able to work in an condition, heartbeat, blood pressure you ask Siri to find an Uber for you
efficient manner to control costs as well and other vital parameters. Doctors and thanks to machine learning that
as get an edge over their competitors. and medical experts can use this Siri and uber both uses which take
1. Financial Services information to analyze the health you home safely. By observing closely
Companies in the financial sector condition of an individual, draw a you will see that machine learning is
are able to identify key insights in pattern from the patient history, and everywhere and that it has become an
financial data as well as prevent any predict the occurrence of any ailments integral part of our lives. We are already
occurrences of financial fraud, with the in the future. The technology also using it, augmenting our lives and
help of machine learning technology. empowers medical experts to analyze moving into future. Machine learning
The technology is also used to identify data to identify trends that facilitate uses algorithms to create immersive,
opportunities for investments and better diagnoses and treatment. interactive experiences between
trade. Usage of cyber surveillance technology and humans. Thus forming
5. Transportation
helps in identifying those individuals or and personalizing a connection to
Based on the travel history and
institutions which are prone to financial technology by automating tasks that are
pattern of traveling across various
risk, and take necessary actions in time typically repetitive or tasks that humans
routes, machine learning can help
to prevent fraud are unable to do such as analyzing the
transportation companies predict
2. Marketing and Sales huge amount of data in forms of text,
potential problems that could arise on
Companies are using machine images or videos. Machine Learning
certain routes, and accordingly advise
learning technology to analyze the will thus continue to shape our world,
their customers to opt for a different
purchase history of their customers connecting us with our friends, family,
route. Transportation firms and delivery
and make personalized product and work.
organizations are increasingly using
recommendations for their next machine learning technology to carry References:-
purchase. This ability to capture, out data analysis and data modeling to [1] Wikipedia : https://en.wikipedia.org/
analyze, and use customer data to make informed decisions and help their wiki/Machine_learning
provide a personalized shopping [2] https://www.outsource2india.com/
customers make smart decisions when
experience is the future of sales and software/articles/machine-learning-
they travel. applications-how-it-works-who-uses-
marketing. Machine Learning is now an it.asp
3. Government important part of our daily lives, [3] https://www.practicalai.io/categorizing-
Government agencies like utilities shaping our society, culture and ever machine-learning-problems/ n

About the Authors

Mr. Chandu D Vaidya working as Asst. Professor at Rajiv on Distributed Operating System. Have Copyright of poster, filled
Gandhi College of Engg. & Research. Have completed patent on IoT application.
post graduation from RTM Nagpur University (RKNEC),
and Graduation from Shivaji University Kolhapur Shubham Atul Uttarwar have completed
(Walchand College of Engg. Sangli). Interested areas of Graduation in Computer Science & Engineering.
research are networking, parallel processing, operating Currently applied for MS and working on
system and cloud computing and published more than Machine Learning.
22 papers in different agencies like Scopus, IEEE,
IJCA etc. Have membership of CSI, and others. Also
published the book on operating System and working

www.csi-india.org
8
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 Technical Trends

A Roadmap for Biological Network Analysis

Nikhila T Suresh, Sreeja Ashok and U. Krishnakumar
Department of Computer Science & IT, Amrita School of Arts and Sciences, Amrita Vishwa Vidyapeetham, Kochi

Significance of Machine learning in Detection of pathways

Biological Network Analysis
The human system is made up of
many interaction networks to carry out
the biological functions in a coordinated
manner. The data intensive nature of Clustering and classification
Gene Finding Algorithms
biology demands computational tools to
analyze and transform the bulk amount
of data into a meaningful prediction
for qualitative treatment of diseases
in health care domain. Machine
learning is an effective platform to
implement expert systems in drug
Motif finding Algorithms
design, diagnosis of genetic diseases,
detection of root cause of disease,
prediction of new diseases etc. Here Prediction of candidate genes
various machine learning approaches
Fig. 2 : Machine learning applications in biological network analysis
for analysis of biological network have
been investigated along with its wide
range of application areas. Biological Networks in analyzing its function. Graph
Human System Protein-Protein Interaction Network approaches have been widely used for
Human system can be considered [PPIN] extracting the network properties.
as a network of components that PPIs are mathematical descriptions Challenges of Modeling Biological
are synchronized with each other to of physical connections between Network
perform various functions by ensuring proteins in the cell. These connections The huge and evolutionary
coordination or rhythm among its are specific, occur at defined regions data booming out from biological
different biological entities (Fig. 1). The and have a particular function or a field through the study of proteins
healthy state of a body is conceptualized specific biological meaning. These (proteome), whole set of information
as the synchronized and uncorrupted models are useful for building scientific about genes (genome), metabolites
processing of these biological modules abstractions and for enhancing the (metabolome), RNA sequence and
or networks. The major categorization biological and biomedical applications.  clinical data generated from history
of these networks are i) Protein- Gene Regulatory Network [GRN] make it impossible to analyze manually
Protein Interaction Networks [PPIN] ii) GRNs have groups of molecular and take a drift from manual based
Gene Regulatory Networks [GRN] and regulators that network in the cell to analysis to computational analysis. For
metabolic Networks [MBN]. manage the  gene  expression levels e.g. Single genome sequence is around
of mRNA and proteins. Thus it can be or greater than 140GB in a hard disk
ORGAN
defined as a network inferred from gene memory size.
NETWORKS

expression data. It represents physical Hence the main challenges are the
biochemical interactions among storage, retrieval and analysis of these
CELLULAR
NETWORKS genes and gene products and is highly complex and heterogeneous biological
phenotype specific. data. Extracting relevant and vital
information which would be beneficial
MOLECULAR Metabolic Network [MBN] in diagnosis of disease, drug designing,
NETWORKS
MBN represents the whole set drug target identification etc needs
of  metabolic  and physical interactions efficient computational solutions.
GENES
that decides the physiological and Machine learning algorithms help in
biochemical properties of a cell. processing these large messy datasets
Understanding the topological to enable intelligent decision-making in
properties of these networks helps a natural way (Fig 2).
Fig.1: Network model of human system

9
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 Technical Trends

Machine learning approaches choice of feature selection method also Prediction of Drug Side Effects”,Cell
Machine learning process includes Systems 2, 209–213, March 23, 2016
determines the classification success
various steps: problem definition, data a2016 Elsevier Inc.
(Pirooznia et al, 2009).
collection, data visualization and data [11] Plaimas et al. BMC Systems Biology
Conclusion 2010, 4:56
validation. Data collection deals with
The data outburst in the field
gathering of data based on features to [12] Emmert-Streib F, Dehmer M, Haibe-
of biological research  necessitates
ensure the quality. Data preparation Kains B. Gene regulatory networks
the  development of systemic and their applications: understanding
involves the randomization to identify
approaches for data pre-processing biological and medical problems in
relationship, data imbalances,
and analysis. The critical advancements terms of networks.  Frontiers in Cell
duplication based on parameters.
in machine learning techniques help and Developmental Biology. 2014; 2:38.
Further improvement can be done doi:10.3389/fcell.2014.00038.
system biology in a great manner. New
by selecting the parameters which
discoveries like drug design, drug target [13] Li YH, Xu JY, Tao L, Li XF, Li S, Zeng X,
ensure optimality to the model is
identification, personalized medicine, et al. (2016) SVM-Prot 2016: A Web-
termed parameter tuning. Particularly, Server for Machine Learning Prediction
effective diagnosis of genetic diseases
the operations such as clustering, of Protein Functional Families from
are taking place in the field of molecular
classification, feature selection and Sequence Irrespective of Similarity.
biology and its one of the promising
motif detection gains importance in PLoS ONE11(8): e0155290. https://doi.
field of not only this era but also in the
biological network analysis. org/10.1371/journal.pone.0155290
coming future.
Cluster is a family of similar [14] Zahiri J, Bozorgmehr JH, Masoudi-
or related entities and clustering is References Nejad A. Computational Prediction of
a process by which the proteins or [1] Muhammad Tahir, Maqsood Hayat Protein–Protein Interaction Networks:
genes of similar functions are grouped ,”Machine learning based identification Algo-rithms and Resources.  Current
of protein–protein interactions using Genomics. 2013;14(6):397-414. doi:10.2
together. Classification is identification
derived features of physiochemical 174/1389202911314060004.
of subgroups on the basis of training properties and evolutionary profiles”,
data and in biological graphs mainly it [15] Sung Hee Park, José A Reyes, David
Artificial Intelligence in Medicine 78
is used to unravel the functionalities R Gilbert, Ji Woong Kim and Sangsoo
(2017) 61–71
Kim,” Prediction of protein-protein
associated with proteins or genes in [2] Supawan Prompramote , Yan Chen interaction types using association
PPIN, GRN respectively. and Yi-Ping Phoebe Chen, “Machine rule based classification”, BMC
In PPI, the clustering algorithms Learning in Bioinformatics” pp 117-147 Bioinformatics200910:36
can be helped in prediction of protein [3] Joseph M Dale, Liviu Popescu, [16] Mohamed Amine Remita1, Ahmed
functions (Li et al. 2016), prediction of Peter D Karp ,”Machine learning Halioui1, Abou Abdallah Malick
interactions (Zahiri et al. 2013), and methods for metabolic pathway , BMC Diouara1, Bruno Daigle1,Golrokh
diagnosis of biomarkers (Wang et al. Bioinformatics 2010, 11:15 Kiani, and Abdoulaye Baniré Diallo,”
2015). Similarly various classification [4] Nitsch et al., BMC Bioinformatics 2010, A machine learning approach for
techniques are used to evaluate 11:460 viral genome classification”, BMC
the druggability of protein-protein Bioinformatics (2017) 18:208
[5] Marenglen Biba et al., ”Using Machine
interactions (Sugaya and Ikeda 2009), Learning Techniques for Modeling and [17] Mehdi Pirooznia1, Jack Y Yang,
prediction of protein interaction (Park Simulation of Metabolic Networks”,2011 Mary Qu Yang, and Youping Deng,” A
et al. 2009), viral genome classification International Conference on Complex, comparative study of different machine
(Remita et al. 2017) etc. The feature Intelligent, and Software Intensive learning methods on microarray gene
Systems expression data”, BMC Genomics 2008,
selection methods in PPI are mainly
9(Suppl 1):S13
used for prediction of essential protein [6] Susmita Datta and Somnath Datta,
(Zhong et al. 2009). ”Methods for evaluating clustering [18] Ni Y, Aghamirzaie D, Elmarakeby H,
algorithms for gene expression data et al. A Machine Learning Approach
In GRN, the clustering facilitates
using a reference set of functional to Predict Gene Regulatory Networks
the analysis of target genes and in Seed Development in Arabidopsis.
classes”, BMC Bioinformatics 2006,
visualization of expression patterns Frontiers in Plant Science. 2016;7:1936.
7:397
(Ni et al. 2016; Pirim et al. 2012). The doi:10.3389/fpls.2016.01936.
classification and feature selection [7] Mordelet and Vert BMC Bioinformatics
2011, 12:389 [19] Piyush B Madhamshettiwar, Stefan
operation in GRN implies GRN R Maetschke, Melissa J Davis,
inference, gene function prediction and [8] Mehdi Pirooznia , Jack Y Yang ,
Antonio Reverter and Mark A Ragan,”
identification of candidate drug targets Mary Qu Yang and Youping Deng, ”A
Gene regulatory network inference:
comparative study of different machine
(Madhamshettiwar et al. 2011; Guo et al. evaluation and application to ovarian
learning methods on microarray gene
2016 ; Bari et al. 2017). cancer allows the prioritization of drug
expression data”,BMC Genomics 2008,
The classifiers in MBN networks targets”, Genome Medicine20124:41
9(Suppl 1):S13
help in the prediction of side effects of [20] Pirim H, Ekşioğlu B, Perkins A, Yüceer
[9] McKinney et al.”Machine Learning for
drugs (Shaked et al. 2016), identifying Ç. Clustering of High Throughput
Detecting Gene-Gene Interactions:”,
essential genes in metabolic pathways Gene Expression Data. Computers &
Appl Bioinformatics. 2006; 5(2): 77–88.
operations research. 2012;39(12):3046-
(Plaimas et al. 2010; Dale et al. 2010). The
[10] Itay Shaked et al, ”Metabolic Network 3061. doi:10.1016/j.cor.2012.03.008.
www.csi-india.org
10
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 Technical Trends

[21] Jiancheng Zhong, Jianxin Wang , Wei [22] Guo et al.,” Gene regulatory network 11:15
Peng, Zhen Zhang, and Min Li,” A inference using PLS-based methods”, [25] Wang et.al, Journal of Cancer 2015, Vol.
Feature Selection Method for Prediction BMC Bioinformatics (2016) 17:545 6
Essential Protein”, Tsinghua Science [23] Shaked et al., 2016, Cell Systems 2, [26] Bari et.al, Nature Reviews Genetics 16,
and Technology, October 2015, 20(5): 209–213 321–332 (2015)
491-499 [24] Dale et al. BMC Bioinformatics 2010, n

About the Authors

Ms. Nikhila T Suresh is a former Assistant Professor at SNGIST Group of Institutions, N Paravur and completed
M.Tech in Network Engineering from Rajagiri School of Science & Tech, Kakkanad and B.Tech from SNGCE,
Kolenchery. Currently pursuing research in Computational Biology at Amrita School of Arts and Sciences,
Amrita University, Kochi.

Dr. Sreeja Ashok is currently working as an Assistant Professor in the Department of Computer Science, Amrita
School of Arts and Sciences, Amrita University, Kochi. She has blend experience in industry and academic
fields. Her industrial experience includes Data Analytics, Project Management, Software Development, and
Quality Assurance (Wipro and Avenir). Her research is broadly in the fields of computational biology, developing
machine-learning algorithms for modeling biological and clinical data; has a strong publication record to her
credit.
Dr. U. Krishnakumar, a Life Member of CSI, is Director, Amrita School of Arts and Sciences and Dean (Arts,
Media and Commerce) of Amrita Vishwa Vidyapeetham University.   He has more than three decades of
experience in teaching and research. Areas of interest include Data Mining, Medical Informatics, Media and
Science Communication. Has published research papers in various national and international journals

Kind Attention :
Prospective Contributors of CSI Communications
Dear Member,
Please note that February 2018 issue of CSI Communications will be the Special issue on Pattern Recognition, whose
Call for Papers (CFP) was already communicated to all CSI members in September 2017. We thank all our CSI members
for the good scholarly response to this special issue.
Only general articles in the categories of Technical Trends, Security Corner and Practitioner Workbench may be submitted
for this issue. Please send your contributions by 20th January, 2018.
The articles should be authored in as original text. Plagiarism is strictly prohibited. Please note that CSI Communications
is a magazine for members at large and not a research journal for publishing full-fledged research papers. Therefore,
we expect articles written at the level of general audience of varied member categories. Equations and mathematical
expressions within articles are not recommended and, if absolutely necessary, should be minimum. Include a brief
biography of four to six lines, indicating CSI Membership no. for each author with high resolution author photograph.
Please send your article in MS-Word format to Editor, Prof. Prashant R. Nair in the email ids [email protected] with cc
to [email protected]
(Issued on the behalf of Editorial Board CSI Communications)

Dr. S. S. Agrawal
Chief Editor

11
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 Technical Trends

Blockchain : A Distributed Ledger

Technology
V. Sai Chandrahaas V. Ravi Sankar**[1]
Indian Institute of Technology(IIT), Madras Dept. of CSE, Gandhi Institute of Technology and Management(GITAM), Hyderabad
Email: [email protected] [1]
Corresponding Author. Email: [email protected]

Introduction in implementing blockchain Banks’ Perspective

Technology experts feel that the technology in the areas of property Banks all over the world have
block chain technology will be a game registration and land records. started experimenting with blockchain
changer and is likely to be adopted by technology. This technology is being
ƒƒ For financial institutions involved in
banks and government departments in adapted for transactions within a bank,
stocks, shares and mutual funds,
the near future. This technology, which between two banks and between banks
blockchain technology can help
is seeing a large scale implementation and financial institutions.
shorten the time involved in trade
abroad, is expected to make huge Instead of recording ledger entries
settlement.
inroads in India too. Blockchain will in a single sheet of paper, a blockchain
facilitate distributed computing in a big ƒƒ In stock market trading, this is created as a distributed database
way. technology can be used in that collects all the transactions
streamlining payments, asset and places them in a block. This
Definition of Blockchain
trading, securities issuance and enables the blockchain to be used as
A blockchain stores list of records
clearing and settlements. a Distributed Ledger Technology(DLT).
in terms of blocks. As the number
of records increase, a new block is ƒƒ This technology can be The DLT enables banks to authenticate
created. There are countless number of implemented in the insurance ownership of assets digitally.
such blocks connected to each other in sector which requires maintenance ƒƒ State Bank of India(SBI), along with
a linear, chronological order. Each block of huge amounts of records of other banks, started working on
is chained to the next block using a premiums, claims and payments. this technology in maintaining Jan
cryptographic signature. The blockchain Dhan accounts.
ƒƒ Industries such as media,
has complete information about diamonds, music and healthcare ƒƒ BNP Paribas, a French investment
different user addresses right from have already started developing bank, has taken up efforts in how
the genesis block to the most recently applications using blockchain blockchain technology can be
completed block. In a blockchain, technology. applied to currency funds and
the data is stored on decentralized order processing.
locations and any change made in the Advantages
ƒƒ Other banks all over the world
records is immediately sent to all the ƒƒ Blockchain technology provides
like – Deutsche bank, Standard
users i.e. each node in the network cost-saving advantage, swift
Chartered, Commonwealth Bank of
gets a copy of the changed block. As the transactions, tamper-proof records
Australia have invested in research
records are stored using cryptography, and absence of a middle party.
and deployment of this technology.
they cannot be modified and tampered
ƒƒ As the technology requires no third
by hackers. Blockchain technology can ƒƒ Companies like TCS and Infosys
party involvement, it is beneficial in
work as an open, online ledger that have already started investing in
cross-border trades, which usually
can record transactions between two blockchain technology to create
take much longer because of time
parties without the help of a third party. core banking platforms for banks.
zone issues.
Hence the need to get authentication
from a third party is eliminated. ƒƒ As the transactions on the network Tech and Financial Companies’
can be done in a decentralized Perspective
Application Areas ƒƒ Citigroup and Goldman Sachs
ƒƒ Blockchain technology has great manner, the technology offers the
ability to create flexible and secure have invested in the development
potential in the public sector area. of payment service engines using
Crucial information like Aadhaar business operations.
blockchain technology.
card numbers, vehicle registration, ƒƒ The electronic ledgers are cheaper
pensions, tax payments, patents to maintain when compared to the ƒƒ IBM’s Hyperledger is an open
and real estate records can be traditional accounting systems. source platform that can be used
registered on the blockchain. to build applications in cloud
ƒƒ The technology helps in minimizing environment using blockchain
ƒƒ Governments of Telangana and the processing delay of stock code.
Karnataka have taken initiatives trading transactions.
www.csi-india.org
12
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 Technical Trends

ƒƒ In December 2015, Microsoft of frictionless commerce. control.

Corporation and ConsenSys
Blockchain Technology and Bitcoins ƒƒ Custom software design and front
announced Ethereum Blockchain and backend programming are
as a Service(EBaaS) on Microsoft’s These days a new crypto-currency
called bitcoin has emerged in the still required to link up the new
cloud computing platform, Azure. blockchain ledgers to the current
The two companies have already world markets. Satoshi Nakamoto,
a cryptographer, created the bitcoin business networks.
started developing an open source,
blockchain based identity system crypto-currency in 2008. Bitcoin is Conclusion
for people, products, apps and a digital currency that allows you to Blockchain technology has started
services. perform peer-to-peer transactions to evolve making its debut in many
without the interference of a third
ƒƒ Google’s Alphabet has started a areas of business and industry. The
party such as banks. Blockchain is the Distributed Ledger Technology is
project to track usage of healthcare underlying technology in developing the
records based blockchain already adopted by several banks and
bitcoins.
technology in the early part of 2017. other companies to manage data and
Issues in implementation of interact with customers and suppliers.
ƒƒ R3[1], an enterprise software firm, Blockchain Technology The technology has caught the attention
is working with over 80 banks, ƒƒ Regulation is the most important of policy makers all over the world and
financial institutions, regulators, aspect in creating an open digital millions of dollars will be invested in the
trade associations, professional environment for commerce and coming years. This technology is behind
services firms and technology financial transactions. Questions the bitcoin and other crypto-currencies.
companies - has developed Corda, like – Who is responsible for It is being endorsed to the government
a distributed ledger platform maintaining and managing the advisory bodies by prominent financial
designed specifically for financial blockchain? Who admits new institutions. Blockchain technology
services. Corda meets the highest participants to the blockchain? is here to stay as it will reduce costs,
standards of the banking industry Who determines which users see increase transparency, store records in
and it is also applicable to any what transactions? Who validates a tamper-proof manner and promote
other commercial scenario. With the transactions? – are yet to be innovation and economic growth.
Corda, participants can record, properly answered.
manage and execute financial References
ƒƒ Banks are not interested in an open [1] https://www.r3.com
agreements in perfect synchrony
with their peers, creating a world source model for their operations [2] https://www.investopedia.com/
as they want to maintain a close terms/b/blockchain.asp

About the Authors

V. Sai Chandrahaas is currently pursuing third year B. Tech in Indian Institute of Technology(IIT), Madras. He is the Head
of Electronics Club at the Centre For Innovation (CFI) in IITM. He also heads the sub module of Navigation, Control and
Communication in IIT Madras’s team in SpaceX’s Hyperloop pod design competition. His areas of interest are Artificial
Intelligence and Machine Learning.

Dr. V Ravi Sankar (CSI Member No. F8000538) is an Associate Professor at Gandhi Institute of Technology and
Management(GITAM), Hyderabad. He received MS(Computer Science) from New Jersey Institute of Technology(NJIT) and
Ph.D from Jawaharlal Nehru Technological University(JNTU), Hyderabad. His research areas are Semantic Web, Data
Mining and Big Data Analytics.

13
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 R esearch  f ront

Collaborative Development of Electro-Optical

Payload Evaluation Software (XSCoPExpress)
Nutan Kumari, Jitendra Sharma and Amit Dave
PCSVD/PCEG/SEDA, Space Applications Centre, ISRO, Ahmedabad – 380015. E-mail: {nutan, jitendrasharma, amitdave}@sac.isro.gov.in

Electro-optical Payload Evaluation Software System, caters to important requirements like Data
Acquisition, Instrument Control, Evaluation of Performance Parameters, Archival and Offline Analysis
of data. It has been implemented as client-server based layered architecture, wherein the processing
and archival is taken care by a pool of high-end servers and data visualization is handled on suitable
clients. Server side development is carried out in C/C++ and Tcl/Tk, Client side development is using
JAVA, and database is around Oracle 11g. It is essential to have a team comprising multiple and different
skill sets for concurrent development of such a critical and complex system, thereby requiring standard
procedures to develop, change and monitor the source code. Version control tools manage the source
code with complete change history and enables synchronization of the code amongst team members.
This paper describes in detail, a need for source code management for complex systems and also, a
practical experience with two most popular version control systems viz. CVS and SVN under LINUX
environment.

Keywords: XSCoPExpress, Subversion, Configuration Management, VCS

I. Introduction
Space Applications Centre is interfaces with checkout sub-systems like perform the payload evaluation tasks.
responsible for realization of Electro- Payload Data Acquisition System (PDAS), At the same time each subsystem
Optical Remote Sensing payloads Spacecraft Interface Simulator (SIS), is developed, tested and maintained
Command and Monitoring System (CMS)
for Earth Observation as well as separately. The system has different
apart from COTS Instrumentation and
Interplanetary missions. Each payload Database etc. sub-system categories like Video
is tested exhaustively in all aspects data acquisition, House Keeping
during development to meet its Data
UTs data acquisition, Data Processing,
performance requirements. Also the Storage
Results++ Data Archival and Retrieval, Results
payloads are different in configurations presentation, Visualization, Optical Test
and hence test methodology varies Bench control etc. All of above call for
from payload to payload, where large very specialized skill sets and expertise.
PDAS XSCoPExpress Software
number of performance parameters Final integrated software product is
are evaluated through different test released to accomplish the tasks.
benches and through the various XSCoPExpress has an agile and
phases of development. Development Instrumentation
accommodative model [2], which
Payload
of generic software on a collaborative Software facilitates interaction and based on the
model, which caters to characterization requirements, necessary changes are
of different types of sensors is a made. These changes fall in following
challenging task. Fig. 1: System Configuration different categories:
XSCoPExpress (UNIX based System 1.  User request based changes:
for Payload Evaluation [1]) caters to the III. Why Version Control Changes in UI, computation,
checkout and evaluation requirements Version control systems (VCS) are algorithms with respect to the
of Electro Optical payloads. Subsequent essential for any form of distributed specific needs. Some changes
sections bring out the reasons for and collaborative development. sometimes apply to all subsequent
collaborative development and how it XSCoPExpress system also falls in this developments.
has been achieved. development category as it has different 2. Technological changes: As
II. System Configuration subsystems which are integrated to underlying systems and/or APIs
XSCoPExpress has both direct and indirect
www.csi-india.org
14
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 R esearch  f ront

change, change at top becomes and Tcl/Tk. Whereas SVN is used as interface for check-in/checkout and
mandatory. the versioning system to share the graphical comparison of version, code
changes of the client side development, branching etc.
3. Improvements: With progress
work which are implemented on JAVA VI. What is SVN?
in payload development, fresh
platform. SVN is used as the client-server
understanding emerges asking for
Reason for the choice was team’s architecture based version control tool
design changes, interface changes,
overall knowledge at the initial phases to manage files and directories, and the
optimization and up-gradation.
of development. With progress and changes made to them over a period of
These aspects need proper version experience, deeper insights were time. This allows sharing the changes,
control to have smooth transition gained with regards to both the version recovering older versions of data or
from older to newer system as well as control systems. examining the history of how data
traceability. This calls for:
Create Authenticate has changed. It is distributed as free
1.  Synchronization of work from Repository Create Users Users software under the Apache License.
different developers;
Modern IDEs like NetBeans and
2. 
Keeping track of changes from Fig. 2a: SVN-Creating Repository Eclipse are equipped with SVN support.
different team members through CLIs are also available for management
change history; tasks of SVN repositories.
3. 
Integrating the final product Repository of software is
• Checkout Project
Check • Merge changes established on the server and users are
with unique version number and • Update local copy • Resolve conflicts
distribution to users; • Make changes
for any changes in
created for authentication using SVN
in code
repository configuration files. New project can be
4. 
Maintaining multiple version, if Changes Commit added (checked-in) to the repository
demands come from different using SVN client after authentication.
category of users or from different Fig. 3. depicts breakup of the
projects at the same time; XSCoPExpress client applications with
Fig. 2b : SVN-Work Flow
5. 
Possibility to revert back the dependencies in the form of common
changes and restore the previous module libraries. Therefore, the
version, if the need arises. V. What is CVS? development needs to be synchronized
CVS is the concurrent versioning for versions and making of final client
IV. Configuration Management Tools system which has client-server application for deployment purpose.
Tools make the work easy to architecture and is used for revision
VII. How SVN is used in collaborative
manage the different versions where control. It has centralized repository,
development of XSCoPExpress
collaborative work is pursued. Many where the server side modules are
tools, in Open Source and Commercial stored. Developer takes the latest code • Access of project from repository
domains, are available for the from the repository for editing the Any number of members can be
configuration control of the software. existing module and then updates the added in the XSCoPExpress team at
Concurrent Version Control (CVS), repository after the modifications. Both any time and they can start the work
Subversion (SVN) and Mercurial are CLI based and GUI tools like Cervisia, immediately. All that is needed is to
most popular amongst the developers Tortoise CVS, Emacs etc. are available check out the respective files and
in Open Source category. ClearCase for CVS. GUI tools provide easy user source code from repository using user
(IBM Rational), Team Foundation
Server (Microsoft), AccuRev SCM (Micro XSCoPE Client Application
Focus International) are some of the
commercially available version control
software.
XSCoPExpress is based on client- Common Module Test Bench Specific Module
server architecture and therefore the
requirements are catered to by server
side and client side components.
The server side requirements are Common requirements like settings, Optical
handled by ASH!Server component. Noise performance Calibration
acquisition, result computation, characterisation
ExpressClient(s) is a set of user visualization, archival etc.
interface applications through which
payload evaluation activities can be
accomplished. Spectral
SNR SNR LTC SWR BBR
CVS is used as the version control Calibration
tools to manage the server side
packages, which are developed in C++ Fig. 3 : XSCoPExpress Client Structure

15
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 R esearch  f ront

ID and password. without any conflict etc. File Types

• Development synchronization XSCoPE development has suitably Type Files LOC LOC per file
Everyone can keep tab on other’s made use of all the above provisions in 126 16635
*.hxx 132.0
changes. Whenever any change in SVN, thereby making it quite easy to pick (3.3%) (2.0%)
project is committed in the repository, the final version of code. Subversion 14256
corresponding files are highlighted in *.doc 6 (0.2%) 2376.0
support integrated with NetBeans IDE (1.8%)
all the local working copies. Developer [3] is used for this purpose. 35 13240
can check the differences in the local *.cpp 378.2
• If required multiple version can (0.9%) (1.6%)
and remote copies and can update their 860 60980
code accordingly. be managed Others 70.9
Testing of multiple electro-optical (22.8%) (7.5%)
• Commit the changes payloads always goes in parallel using
Once the development work is Above statistical information for a
same XSCoPE application. They all
over and code is fully tested, it can be will have different configurations project gives an idea about various
submitted to the repository. Developers types of source files, their LOC and an
and to characterize them on ground,
first update their project with the average LOC per file.
requirements may vary from one to
repository. There are two possible
another. To handle such a scenario Following table contains list of Top
cases:
different versions of XSCoPE may be
1. Project will be updated successfully 10 source code files for a project. It
needed at the same time. At present,
2. 
Conflict may occur. In this case, this case is handled without any extra contains mix of .java and .cxx files.
project problems are resolved first. effort with the help of SVN. Largest Files (TOP 10)
After updating successfully, project VIII. Statistics of XSCoPE-Repository File LOC
is ready to be committed. Developer has XSCoPE is a complex system and ExpressCommon/…/GenericApp.java 18394
the choice to either commit the project has many components. Challenges
fully with all changes or partially with ExpressCommon/…/EMGenericApp.
involved in development and java
9397
only subset of modules using registered maintenance of such critical system
user ID and password. Comments ExpressCommon/…/
developed over a period of time is self- ImageBasedExplorer.java
9083
about the changes are also given explained from the statistics of XSCoPE
in the marked area to keep track of ArsenalDev/…/Aeu.cxx 6944
repository generated using the StatSVN
modification in future. tool [4]. ArsenalDev/…/configurations.xml 6098
• To keep track of changes ArsenalDev/…/Beu.cxx 5757
• Development Statistics
Team leader assigns the job to all Report Generated: 2017-08-31 15:04 ExpressCommon/…/
5607
XSCoPE team members and can check Head revision: 556 AcqSettingsDialog.java
the project history at any time to see Report Period: 2013-04-11 to 2017-08-03 ArsenalDev/…/bbrc.cxx 5349
the development progress. If the short Total Files:3768 ExpressCommon/…/ImgPanel3D.java 4345
detail provided by developer is not Total Lines of Code:811676
Developers:7 ArsenalDev/…/lvdr.cxx 3568
sufficient to pin point the exact changes,
he can compare any two versions to find • File Statistics
Top 10 Files With Most Revisions
the modifications in the code. What, File Types
where and why the changes, can be File Revisions
Type Files LOC LOC per file
answered easily. ExpressCommon/…/GenericApp.java 138
3768 811676
• To revert back the modifications Totals 215.4 ExpressCommon/…/
(100.0%) (100.0%) 84
If the team leader is not satisfied AcqSettingsDialog.java
420 289903
*.c 690.2
with the changes, the changes can be (11.1%) (35.7%) ExpressCommon/…/
37
reverted back and the previous version 471 179038 PanelViewerXResult.java
*.java 380.1
of project can be restored. (12.5%) (22.1%) ExpressSNR/…/ExpressSNR.java 28
• Easy to distribute 127 121101 ExpressCommon/…GenrateXYPlot.java 23
*.cxx 953.5
Before handing over the final (3.4%) (14.9%)
ExpressCommon/…/ASatInfo.java 23
code to the user, following needs to be 141 37184
*.xml 263.7 ExpressCommon/…/
(3.7%) (4.6%) 20
verified: PanelViewerXResult.form
a. Which is the latest version? 99 32035
*.form 323.5 ExpressCommon/…/
(2.6%) (3.9%) 18
b. 
Which system contains the final ImageBasedExplorer.java
368 24360
code? *.h 66.1 ExpressSNR/…/SNRSettingsDialog.
(9.8%) (3.0%) 17
18 22944 java
c. Whether that code contains all the *.xbm 1274.6
changes from all the developers (0.5%) (2.8%) ExpressInstaller/…/ASH_Home.java 16

www.csi-india.org
16
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 R esearch  f ront

•  Developer
Developer Activity
Activity And application
finally, not manycoding
tools areisbeing
in SVN repository.
developed ƒƒ It is easy to use and install --
SoCVS
around there wasNetBeans
anymore. a needIDE
to has
switch
stoppedfrom CVS especially on client machines.
CVSto Subversion.
support since version 7.0. Whereas, SVN has a
ƒƒ Network security: Subversion can
pretty large Main challenge
application to transfer
base on any system including the data
use https or svn+ssh which make
IDEsfrom CVS to SVN repository was not to
like NetBeans.
it safe to use remotely. ClearCase
just transfer the latest copy of data, but
has to be used via VPN to be truly
all X. the development
Migration from CVS history
to SVN as well.
secure and some setups of VPN
cvs2svn [5] is a tool for(CVS)
migrating a
Migration from older version control system filter ClearCase packets out.
was CVS repository
inevitable because of to the Subversion
following important or Git. The
main design goals are robustness and
reasons:
ƒƒ Subversion is much faster and
1. To100% data system
have a uniform preservation.
for all applications Conversion
leaner.
2. Totooltakecvs2svn has ofbeen
the advantages new used
features to
of migrate ƒƒ Subversion makes a clear
the repository
Subversion over CVS. to SVN repository without distinction between the client and
losing the latest history. It converts server. An administrator only has to
IX. Comparison
IX. of CVS and
Comparison of CVS and SVN SVN all revisions,
As discussed earlier, the serverbranches,
code of XSCoPEtags, is log worry about the Subversion server
CVS is based on RCS files of beingmessages, managed using CVS authorrepositorynames, and commit
and the client installation and not how the clients
CVS is based on RCS files of versions control. Each
versions control. Each file connected to application dates. codingcvs2svn also deals
is in SVN repository. with many
So there was are setup.
file connected to CVS is an ordinary file containing
CVS is an ordinary file containing some a CVS
need to oddities
switch from CVSandto is highly
Subversion. configurable.
some additional information. Thus, it works on file-
additional information. Thus, it works XI. Challenges Faced Conclusion
by-file basis. Being simple, with CVS, data loss is
on file-by-file basis. Being simple, with Main1.  During commit,
challenge to transfer the data from CVS to SVN
if some For collaborative development,
minimal, repository corruption can be easily recovered
CVS, data loss is minimal, repository repository was not to just transfer the latest copy of
interruption happens and commit version control system is a must. Lack
by correcting RCS files, if necessary.
corruption can be easily recovered by data, but all event the development history as well.
is aborted in the middle, files of proper tools is one of the reasons
Basis of SVN is a relational database BerkleyDB
correcting RCS files, if necessary. get locked. [7] for poorly engineered product and
composed of a set of binary files. SVN tracks the entire
Basis of SVN is a relational cvs2svn is a tool for migrating a CVS repository
[5]
delayed delivery. For a centralized
commit as a new revision. SVN’s base being a set of 2.  Two or Git. developers have either requirement like XSCoPExpress, where
database BerkleyDB composed of a set to Subversion The main design goals are
binary files, allows more concurrency and is faster in changed the same portions of code the development is confined and non-
ofterms
binary files. SVN tracks the entire robustness and 100% data preservation. Conversion
of speed of operation. or modified the same variables distributed, both CVS and SVN are
commit as a new revision. SVN’s base tool cvs2svn has been used to migrate the repository
and committed
without losing the the latest changes in the good. Mercurial and Git are based on
being a set of binary files, allows more to SVN repository history. It
Atomic commit operation in concurrency environment repository. distributed repositories and not bundled
concurrency and is faster in terms of converts all revisions, branches, tags, log messages,
is a problem in CVS. Whereas in SVN this problem
speed of operation. 
author Conflict
names, and commit willdates. occur
cvs2svnin alsothis
deals situation with standard Linux distribution. SVN
does not crop up.
Atomic commit operation in with many CVS and only one change will be
oddities and is highly configurable. is more robust, reliable, newer and
concurrency environment is repository,
a problem accepted
XI. finally.
Challenges Faced So the other available out-of-the-box on standard
CVS allows rollback of any commit in the
ineven
CVS. Whereas in SVN this problem developer has to get the final copy
1. During commit, if some interruption happens and Linux distribution. Hence, SVN turns
though this may require some time, as each file
does not crop up. from repository and repeat the
commit event is aborted in the middle, files get out to be the best choice for limited
is processed independently. SVN does not allow
CVS
rollback allows rollback of any commit
of commit.
locked.
work once again. projects.
in the repository, even though this 3.  Subversion does not know about Acknowledgment
may require some time, as each file is the conflict free merges which We sincerely acknowledge constant
processed independently. SVN does not makes it impossible to track who motivation and encouragement from
allow rollback of commit. has changed which code. So when Mr. Ashish Mishra, Division Head, and
And finally, not many tools are two developers are working on Mr. D.R. Goswami, Group Director,
being developed around CVS anymore. same set of code, then in future it Payload Checkout Evaluation Group.
NetBeans IDE has stopped CVS support will be difficult for a team leader to We also thank Mr. S.S. Sarkar, Deputy
since version 7.0. Whereas, SVN has track the changes if changes are Director, SEDA for providing opportunity
a pretty large application base on any not properly commented. to work on Payload Evaluation System.
system including IDEs like NetBeans.
References
X. Migration from CVS to SVN XII. Comparison of SVN with other [1] Amit Dave, Jitendra Sharma and Anil
Migration from older version commercial tools Sukheja; A System for Concurrent
control system (CVS) was inevitable As mentioned in section IV, many Evaluation of Multiple IRS Cameras;
because of the following important tools, in Open Source and Commercial ISRO CONFERENCE ON COMPUTER
domains, are available for the TECHNOLOGY FOR SPACE (CTS 2011);
reasons:
configuration control of the software. Bangalore; August 2011
1. To have a uniform system for all
applications One comparison chart [6] between [2] XSCoPExpress Client Application
Subversion and Commercial Tool Software Design Document ID: SAC/
2.  To take the advantages of new ClearCase is shown below: PCSVD/SW/XSCOPE/01/2015/1.2
features of Subversion over CVS. Advantages of Subversion over [3] Using Subversion Support in NetBeans
As discussed earlier, the server ClearCase: IDE, https://netbeans.org/kb/docs/ide/
code of XSCoPE is being managed ƒƒ SVN is available on a lot more subversion.html
using CVS repository and the client platforms than ClearCase. [4] http://statsvn.org – Tool to retrieve

17
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 R esearch  f ront

information from subversion repository. [6] [6] http://pti.co.il/clearcase_vs_ of Reasons for Delay in Software
subversion.html Development; IEEE Transactions On
[5] [5] http://cvs2svn.tigris.org – Website
Software Engineering, VOL. 17, NO. 6,
for Open Source Software Engineering [7] [7] Michiel van Genuchten; Why is
JUNE 1991
Tools. Software Late? An Empirical Study
n

About the Authors

Mr. Amit Dave is Sci/Engr “SG” – B.E. Computer Science M. S. University, Vadodara. He is associated with development of
software for ground checkout of various EO payloads. Developed device drivers for high-speed data acquisition card, payload
checkout applications, algorithms, instrument control, automation, database, web based software for various EO payloads
for earth-observations, inter-planetary missions and airborne sensors.

Mr. Jitendra Sharma Sci/Engr – “SF”. M.Tech Computer Sc. And Technology (IIT Roorkee), B.E. Electronics (MMMEC
Gorakhpur). He is currently working on High performance computing, GPGPU, Distributed Computing, Micro controllers,
Airborne Imaging Systems.

Ms. Nutan Kumari, Sci/Engr “SD”– B.Tech Comp Sc. (BVP Deemed University, Pune). Presently working on
payload checkout evaluation system, automation of different test benches, High performance computing,
GPGPU, Airborne Imaging Systems.

IEEE MGA Leadership Award to

Mr. H R Mohan
CSI Fellow & CSI Past President

CSI Communications congratulates Mr. H R Mohan, CSI Fellow and CSI Past President
for having been selected as a 2017 recipient of the IEEE MGA Leadership Award. He is
currently the Vice Chairman of IEEE Madras Section, the largest Section in India. The
official communication to Mr. Mohan from IEEE states:
This is a very prestigious award, and we are pleased to honour you for the outstanding
contributions you have made to numerous IEEE activities and the example you set in
carrying forward the goals and objectives of the IEEE MGA Board.
Please accept our congratulations on being recognized for this most deserving honour.
The citation states: For dedicated and sustained commitment to member engagement
through various programs, student activities and information sharing.

www.csi-india.org
18
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 R esearch  f ront

Text Detection and Recognition in

Natural Scene Images
Sharath Kumar Y H Puneeth P
Dept. of Information Science & Engg, Maharaja Inst. of Tech., Dept. of Information Science & Engg, Maharaja Inst. of Tech.,
Belawadi, Srirangapatna Tq, Mandya, Karnataka, India Belawadi, Srirangapatna Tq, Mandya, Karnataka, India
[email protected] [email protected]

1. Introduction
In context of scanned machine Issues having an impact on image features are extracted from scene text.
written text-detection and recognition is processing are limited to sensor noise, Finally, the character identification and
considered a solved problem. There are skewed documents and inherent recognition is done using OCR. Figure 1
several commercial and freely available degradations to the hand written shows the natural scene images which
systems [25], achieving recognition document. Based on the classical containing text.
performance of 99%. Text detection and acquisition method, the optical 2. Background Work
recognition in natural scene images character recognition (OCR) systems To understand the overall value
is, however, a challenging, unsolved have been designed for many years of text detection and recognition
computer vision problem. Scene text to reach a high level of recognition approaches, it is useful to provide
has complex background, image blur, with constrained documents. The background information about the
partially occluded text, variations in recent explosion of handheld imaging underlying problems, applications and
font-styles, image noise and varying devices (HIDs), i.e. digital cameras, technical challenges.
illumination. Commercial systems do standalone or embedded in cellular
not work in these setting. The end-to- phones or personal digital assistants 2.1 Text in Imagery
end scene text recognition problem (PDAs). Research on document image Graphic text and scene text are
is divided into a text-detection and analysis [24] entered a new era where considered two basic classes of text,
text recognition task. Text-detection breakthroughs are required, traditional where the former refers to machine
is a preprocessing step for the text- document analysis systems fail against print text overlaid graphically and the
recognition task. The text detector has this new and promising acquisition latter refers to text on objects, captured
to locate words in natural scene images. mode. in its native environment. Graphic text
Hence the detector has to distinguish In this work, we aim to develop novel is usually machine printed, found in
text locations from non-text locations in approach for detection andrecognition captions, subtitles and annotations in
an image. The text-recognizer predicts of natural scene text. Initially the real video and born-digital images on the
a word shown in a cropped image patch world images have taken. The various web and in email [37].
which is retrieved by the detector.
Scene text identification and
recognition is used widely nowadays,
natural scene images, especially in
modern urban environment for example
(posters, product tags, guide posts,
electronic sign boards, shop boards
and name plates etc.) consists of wide
range of data which are very beneficial
to a variety of applications example:
navigation system, target geo-location
product search, data indexing, image
understanding. Scene text consists of
backgrounds that are not necessarily
clean and white, and more complex ones
make text extraction from background
difficult. Moreover scene text such
as that seen in advertisements may
include artistic fonts Fig. 1: Natural scene images containing text

19
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 R esearch  f ront

proposed a method called strokelets

it’s a multi scale representation which
capture the substructures of characters
at different granularities with respect
to the part prototype. Bag of strokelets
feature is used as a descriptor for image
regions that contains text strings for
feature extraction and finally a random
(a) (b) (c) forest classifier is used for classification.
Fig. 2 : Text in imagery. (a) Video graphical text. (b) Point-and-shoot scene text. (c) KAIST et al., [4] proposed a system to
Incidental scene text detect the characters using conditional
Scene text, however, includes text iOS and Android, which can instantly random fields (CRF). The CRF is used to
on signs, packages and clothing in recognize and translate text into form super pixel and passes them to soft
natural scenes, and is more likely to another language [34]. OCR engine for detection of character.
include handwritten material [38]. Most According to the World Health Neumann, L et al., [5] proposed a
recent research has focused on scene Organization, approximately 39 million method for text localization and
text, and, to portray the challenges legally blind and 285 million visually recognition by hypotheses verification
more accurately, it helps to further impaired people live in the world. framework. Here segmentation is done
distinguish between images where Developing personal text-to-speech by maximally stable external region
the primary purpose of the image is to devices assists them in understanding (MEER) and multi-class support vector
capture text and images where the text grocery signs, product and (SVM) used for recognition. A Mishra et
is embedded in the scene. Although a pharmaceutical labels, and currency al., [6] present a framework that exploits
continuum exists between the two, we ATM instructions [35]. both bottom-up and top-down cues.
refer to the former as point-and-shoot Industrial automation. Recognizing The bottom-up cues are derived from
text and the latter as incidental text, as text on packages, containers, houses, individual character detection from the
shown in Fig. 2. and maps has broad applications related image, top-down cues obtained from a
2.2 Applications to industrial automation. For example, lexicon-based prior.
Over the past two decades, there recognition of addresses on envelopes Khaoula E et al.,[7] proposed a
have been numerous text related is applied in mail sorting systems. novel method to recognize scene text.
applications for both images and video, Automatic identification of container It uses multi-scale scanning scheme
which canbe broadly categorized as numbers improves logistics efficiency. using windows with non-linear borders.
multimedia retrieval, visual input and Recognition of house numbers and text Convolutional neural networks is
access, and industrial automation. in maps benefits automatic geocoding designed and trained to classify each
Multimedia retrieval. Text in web systems [36]. window in order to recognize valid
images is relevant to the content of 3. Literature Survey characters and identify non-valid ones.
the web pages. Video captions usually In this section, we provide a brief Xin He et al., [8] proposed a neural
annotate information about where, when review on existing work related to our network-based scene text recognition
and how of the happening events [29], work. system. Sliding window is used over each
[30]. Recognizing text and extracting Baoguang Shi et al.,[1] proposed connected component for segmentation
keywords in such multimedia resources a deep learning based algorithm called purpose and MQDF classifier used for
enhances multimedia retrieval. Multi-stage spatially-sensitive pooling classification. B Shi et al.,[9] proposed
Visual input and access. The network (MSPN). The convolutional a method called robust text recognizer
expansion of mobile devices containing neural networks(CNN) and spatially- with Automatic Recitation (RARE) which
digital cameras has made imaging sensitive multi-stage pooling layerare is a specially-designed with deep neural
devices widely available. With an used for feature extraction and script network. An image is firstly rectified
embedded module, mobile devices identification. Deepak Kumar et al., [2] via a predicted Thin-Plate-Spline (TPS)
automatically input name cards, have proposed effective methods for transformation into a more “readable”
whiteboards and slide presentations segmentation and recognition;itsplit image. Fiducial points are identified by
[31], [32]. Without being forced to the word image into individual color, localization network. By using a grid
input by keyboard, users feel more gray and lightness planes and later generator new set of fiducial points are
comfortable and work more efficiently. enhance the contrast in each of these identified by using Spatial Transformer
Signs in natural scenes carry significant planes independently by a power-law Network (STN) and a Sequence
information. Automatic sign recognition transform. The discrimination factor is Recognition Network (SRN).Xinhao
and translation systems enable users used for feature extraction. The plane Liu et al., [10]have proposed a highly
to overcome language barriers [33]. that has maximum discrimination accurate character recognition model
Carnegie Mellon University developed factor is selected for segmentation; called Convolutional Neural Network
an early PDA-based sign recognizer finallybinarized words are used for (CNN). Here classification is done by
[33], and recent platforms include recognition. Xiang Bai et al.,[3]have Multi-layer perceptron which map’s the
www.csi-india.org
20
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 R esearch  f ront

input to the n different convolutional are used for encoding of words. P non-uniform illumination and noise.
layers. Later Sliding window method Kumar Singh et al., [17] proposed Connected component extraction is
is used for cropped image recognition. page level script identification for hand unable to handle broken strokes and
XiaohangRen et al.,[11] have proposed written documents. The gray level co- connected characters. Classification
a CNN based scene text detection occurrence matrix (GLCM) features of natural scene text was not done
algorithm with a new text region and multi-layer perceptron (MLP) effectively, non-horizontal text detection
extractor. Candidate text region classifier performs the best in page have been a major challenge. Sliding
extraction is done by Maximally Stable level script identification. Xwang et al., window produces plenty of false alarms
Extremal Region (MSER), which improve [18] presented an efficient text detection due to significant variation in character
the independency and completeness of method with multi-channel connected aspect ratio.
the extracted candidate text regions component. Shivananda V.S et al., [19] It is also noticed that no work has
and detection is done by CNN. Deepak presented a hybrid methodology which been attempted towards estimation of
Kumar et al., [12] proposed a midline extracts multilingual text from natural multiple skews in multilingual scene
analysis segmentation technique by scene image. Edge features extracted text. This poses new challenges in the
choosing the labels of middle row using Contourlet transform and field of document image analysis and
of the image as a sub-image. Then, classification based SVM classifier. has motivated us to take up this research
the labels from this segmented sub- A Mishra et al., [20] present a work. The gap between the technical
image are used to propagate labels framework that uses a higher order prior status and the required performance
to other pixels in the image. Bayesian computed from an English dictionary to indicates that text detection and
classification and Max-flow methods recognize a word, which may or may not recognition remain unsolved problems.
have been independently used for be a part of the dictionary. Character While great progress has been made
label propagation and OMNI OCR for detection is by sliding window detector there are still numerous research
recognition. Jay H. Bosamiya et al., and recognition of words based on opportunities.
[13] have proposed and introduced an graph construction. Chunheng Wang End-to-end recognition. Compared
approach for character segmentation et al., [21] propose a novel scene with the performance of OCR on clean
which combines the advantages of text detection approach using graph documents, end-to-end text recognition
connected component analysis, stroke model. The Maximally Stable Extremal performance is still far behind.
width analysis and graph cuts. Regions (MSERs) is used to incorporate Improvement will come not only from
J VivekVeeriah et al., [14] proposed various information sources into one stronger character recognition models,
a system to integrate with augmented framework. Segmentation is done by but also from well-designed information
reality technique such as translation MSER’sLabeling text and non-text sharing, feedback and optimization
from one language to another language. regions and this labeling is used for the strategies. Recently developed large
This method is capable of locating text graph construction. Baoguang Shi et scale deep learning has substantially
written in different languages obtained al., [22] proposed a methodology which improved character classification
from videos. The method uses otsu and combines the deep features and mid- performance by learning hierarchical
MSER for segmentation and cascade level representations into a globally multi-scale representations [39], [40].
of N filters for text region filtering. trainable deep model. .K. C. Kim Shi The integration of deep learning with
Shangkwantian et al., [15] proposed a et al., [23] proposed a method that optimized segmentation, recognition
robust text segmentation method that extracts text regions in natural scene and high order language models could
employs both Markov Random Field images using low-level and high level further boost the performance. Open
(MRF) and graph cut algorithms. Stroke image features. vocabulary recognition. Existing word
feature transform is adopted to robustly 4. Future Nchanment spotting approaches with small lexicons
identify text seeds and text edges, In this section, we provide a brief report encouraging performance with
labeling of 1 for text pixels and 0’s for review on existing work related to our the ICDAR benchmarks. However, the
background pixels using gibbsenergy work. utility of these approaches is limited
E for classification. Max Jaderberg et From the literature survey, we because general lexicons are unlikely
al., [15] proposed encoding schemes found the existing methods with to contain the proper nouns and
like 90k-way dictionary encoding, drawbacks. Many of the work proposed other words that appear in scene or
character sequence encoding and bag- are deals with less dataset and graphical text imagery. To overcome the
of-N-grams encoding. A deep neural inefficient scene text images. Majority limitation, it is useful to incorporate an
network model and data produced by of the proposed systems have used open vocabulary, i.e., large scale web-
a synthetic text generation engines binarization, which is very sensitive to based language information [41].

21
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
R esearch  f ront

ICDAR’11 Graphic text dataset

ICDAR’11 Scene text dataset

MSRA-II Scene Text Dataset

NEOCR Scene Text Dataset

MSAR-I Graphic Text Data Set

Figure 3. Sample images of text from the ICDAR’11, MSRA-I, MSRA-II, SVT, and NECOR datasets.

www.csi-india.org
22
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 R esearch  f ront

characteristic. From the observation,

text, which exists in the natural scenes,
usually appears on color uniformed
surface (e.g. signboard background).
The basic idea of our methodology is to
search the signboard area firstly; text is
then extracted from the signboard area.
So, in this section, we first segment
a scene image into homogeneous
Fig. 4 : Word examples from dataset ICDAR’11 regions. A mathematical morphological
operator, the Toggle Mapping, is applied
It is also useful to develop text say?” on a grayscale image for segmentation
approximation methods to efficiently according to the intensities of pixels.
Text Detection via Edge cue:
use large scale language information Scene text detection aims to locate Skew Detection:
for recognition. Processing incidental the position of texts in different scenes, Hough transform algorithm
text. Incidental text suffers from image e.g. guidepoints, store marks and is used for the skew detection.
degradation, distortions, font variations warning signs, which is one of the most The Hough transform is a feature
and cluttered backgrounds. Many important steps for end-to-end scene extraction technique used in image
approaches can tackle single issues, yet text recognition. Inspired by the fact that analysis,computer vision,and digital
few approaches process a combination edge is one of the most important cues image processing. The purpose of
of them. To address the general to distinguish texts from background, technique is to find imperfect instances
problem of incidental text detection and our methodology consists of two main of objects within a certain class of
recognition, improved invariant features stages, i.e. candidate character region shapes of voting process. this voting
must be designed or learned, state-of- extraction and region classification. For process is carried out in a parameter
the-art enhancement and rectification the stage of candidate region extraction, space, from which object candidates
methods must be integrated, and new the method extracts the edges of an are obtained as local maxima in so
sensors must be applied. input image, which will be used to called accumulator space that is
Processing multilingual text. Text determine the possible positions where explicitly constructed by the algorithm
from various languages shows different the texts will be appeared and then for computing the hough transform.
characteristics. Recognition of text slides multi-scales windows around The classical Hough transform was
from East Asia countries, e.g., China, the edge connected regions. For the concerned with the identification of
Japan and India, was considered an stage of region classification, the lines in the image ,but later the Hough
extremely difficult problem due to the method extracts and combines high transform has been extended to
large number of character classes, discriminable hand-designed low level identify the position of arbitrary shapes
complicated character structures, features and CNN based high-level especially circles or ellipse.
the similarity among characters, and features for region classification. Text Recognition via CNN classifier
the variability of fonts. Using a single
Text Extraction via Homogeneous and WFST-based word labeling
method with fixed parameters to
region segmentation: The word recognition task
recognize text from all the languages
Segmentation is the process of consists of identifying characters and
remains difficult to achieve.
partitioning a digital image into multiple recognizing them as text format from
5. Methodology regions of pixels. Each of the pixels in a a given image patch. Our method
Text Detection and Localization: region are similar with respect to some uses the CNN-based word labelling
“Are they any image in the text and characteristic or computed property, approach for efficient and accurate text
where is the text?” such as color, intensity, or texture. recognition. We are going to develop an
Text Extraction: “How can segment Adjacent regions are significantly accurate character recognition model
text from the background?” different with respect to the same based on Fractional Max Pooling (FMP)
Text Recognition: “What does the structure and Network in Network (NIN)
module. Then for post processing, we
Input image
Text Extraction by develop a WFST based word labelling
Text Detection Classification Homogeneous region approach which encodes a mapping
And Localization segmentation from the CNN labels to desired lexicon
word. The benefit of the WFST based
representation is that it is not only
efficiently handle the ambiguous set
Out processed Skew Detection
Text Recognition of CNN labels, but also very flexible to
text and correction
incorporate with high order lexicon and
Fig. 5 : System Diagram of Natural Scene Detection Recognition language models.

23
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 R esearch  f ront

6. Conclusion Tong University, Shanghai, China. Borovikov American Management

In this work we briefly discuss the [12] Deepak Kumar M N Anil Prasad A Systems, Inc. 4831 Walden Lane,
Application and Challenges of Natural G Ramakrishnan MILE Laboratory Lanham, MD 20706
Dept. of EE Indian Institute of Science [26] Weilin Huang et al. Robust Scene Text
Scene images. In addition, we also
Bangalore, INDIA [email protected]. Detection with Convolution Neural
review the various work proposed by ernet.in. Network Induced MSER Trees The
the researchers across the world on [13] Jay H. Bosamiya, Palash Agrawal, Chinese University of Hong Kong,
natural scene and we also identified Partha Pratim Roy, R. Balasubramanian China,2014
the drawbacks of the work proposed. Department of Computer Science [27] A. Shahab, F. Shafait, and A. Dengel.
Based on the drawbacks we drafted the and Engineering Indian Institute of ICDAR 2011 Robust Reading
objectives for the future work. Finally, Technology Roorkee, India 247667 Competition Challenge 2: Reading Text
[14] J Vivek Veeriah Dept. of ECE Coimbatore in Scene Images. In Proceedings of the
the proposed methodology for text
Institute of Technology , Deepak Kumar International Conference on Document
detection and recognition is discussed MILE Laboratory Dept. of EE Indian Analysis and Recognition, pages 1491–
briefly. Institute of Science Bangalore, India. 1496. IEEE Computer Society, 2011.
References K R Ramakrishnan CVAI Laboratory [28] L. P. Sosa, S. M. Lucas, A. Panaretos,
[1] Baoguang Shi*, Cong Yao*, Chengquan Dept. of EE Indian Institute of Science L. Sosa, A. Tang, S. Wong, and R.
Zhang*, XiaoweiGuot, FeiyueHuangt, Bangalore, India. Young. ICDAR 2003 Robust Reading
Xiang Bai* *School of EIC, Huazhong [15] Shangkwantian, shijianlu, tbolonsu and Competitions. In Proceedings of the
University of Science and Technology, chew lim tan Department of Computer International Conference on Document
Wuhan, P.R. China 430074 tTecent, Science, School of Computing, National Analysis and Recognition, pages 682–
Shanghai, P.R. China 200233 University of Singapore Computing 1, 13 687. IEEE Press, 2003.
[2] Deepak Kumar, M N Anil Prasad and A Computing Drive, Singapore 117417 [29] Y. Zhong, H. J. Zhang, and A. K. Jain,
G Ramakrishnan Medical Intelligence [16] Max Jaderberg Karen Simonyan, “Automatic caption localization in
and Language Engineering Laboratory Andrea Vedaldi Andrew Zisserman, compressed video,” IEEE Trans. Pattern
Department of Electrical Engineering Visual Geometry Group, University of Anal. Mach. Intell., vol. 22, no. 4, pp.
Indian Institute of Science, Bangalore, Oxford 385–392, Apr. 2000.
[17] Pawan Kumar Singh#, Santu Kumar [30] Q. Ye, Q. Huang, W. Gao and D. Zhao,
INDIA-560012
Dalal, Ram Sarkar, MitaNasipuri “Fast and robust text detection in
[3] Xiang Bai, Senior Member, IEEE, Cong
Department of Computer Science images and video frames,” Image Vis.
Yao, and Wenyu Liu, Senior Member,
and Engineering, Jadvapur University, Comput.,vol. 23, pp. 565–576, 2005.
IEEE
Kolkata-70032, India [31] J. Liang, D. Doermann, and H. Li,
[4] KAIST AIPR. http://ai.kaist.ac.kr/home/.
[18] Xiaobingwang, yanohong song, “Camera-based analysis of text and
[5] Neumann, L. and Matas, J., “A Method
ywenlinzhang. Institute of artificial documents: A survey,” Int. J. Doc. Anal.
for Text Localization and Recognition in
intelligence and robotics. Recognit., vol. 7,pp. 84–104, 2005.
Real-World Images,” Proc. 10th Asian
[19] Shivananda V. Seeri, S. Hiremath KLE [32] L. Lin and C. L. Tan, “Text extraction
Conf. on Computer Vision (ACCV), 770-
Technological University, BVBCET fromname cards using neural network,”
783, (2010).
Campus, Hubballi, Karnataka, J. D. in Proc. Int. Joint Conf. Neural Netw.,
[6] Mishra, A., Alahari, K. and Jawahar,
Pujari2, P.S.D.M. College of Engg. & 2005, pp. 1818–1823.
C.V., “Top-Down and Bottom-Up Cues Tech., Dharwad, Karnataka [33] X. Chen, J. Yang, J. Zhang, and A. Waibel,
for Scene Text Recognition,” Proc. IEEE [20] Anand Mishra1, Karteek Alahari2, “Automatic detection and recognition of
Conf. on Computer Vision and Pattern C.V. Jawahar1 ,1: CVIT IIIT Hyderabad signs from natural scenes,” IEEE Trans.
Recognition (CVPR), (2012). Hyderabad, India 2: INRIA - WILLOW ImageProcess., vol. 13, no. 1, pp. 87–99,
[7] Khaoula Elagouni Orange Labs R&D ENS Paris, France Jan. 2004
Cesson-S´evign´e, Christophe Garcia [21] Chunheng Wang⇑, Baihua Xiao, Yang [34] A. Bissacco, M. Cummins, Y. Netzer,
LIRIS, Insa de Lyon Villeurbanne, Zhang, Song Gao State Key Laboratory and H. Neven, “Photo OCR: Reading
Franck Mamalet Orange Labs R&D of Management and Control for Complex text in uncontrolled conditions,” in Proc.
Cesson-S´evign´e, Pascale S´ebillot Systems Institute of Automation, IEEE Int. Conf.Comput. Vis., 2013, pp.
IRISA, Insa de Rennes Rennes, Chinese Academy of Sciences, Beijing 785–992.
[8] Xin He, Yi-Chao Wu, Kai Chen, Fei Yin, 100190, China [35] X. Liu and D. Doermann, “A camera
Cheng-Lin Liu National Laboratory [22] Baoguang Shi, Xiang Bain, Cong Yao phone based currency reader for
of Pattern Recognition Institute of School of Electronic Information and the visually impaired,” in Proc. ACM
Automation of Chinese Academy of Communications, Huazhong University SIGACCESSConf. Comput. Accessibility,
Sciences, Beijing 100190, China of Science and Technology, Wuhan 2008, pp. 305–306.
[9] Baoguang Shi, Xinggang Wang, 430074, China [36] P. Sermanet, S. Chintala, and Y. LeCun,
PengyuanLyu, Cong Yao, Xiang Bai* [23] K. C. Kim1, H. R. Byun1, Y. J. Song2, Y. “Convolutional neural networks applied
School of Electronic Information and W. Choi2, S. Y. Chi3, K. K. Kim3, Y. K. to house numbers digit classification,”in
Communications Huazhong University Chung3 1Dept. of Computer Science, Proc. IEEE Int. Conf. Pattern Recognit.,
of Science and Technology Yonsei University, Korea 2Dept. of 2012, vol. 4, pp. 3288–3291.
[10] Xinhao Liu, Takahito Kawanishi, Computer Science, Sookmyung [37] D. Karatzas, S. Robles Mestre, J. Mas,
Xiaomeng Wu Kunio Kashino NTT Women’s University, Korea and F. Nourbakhsh, “ICDAR 2011
Corporation, Kanagawa, Japan 243- [24] ya ro s l a v v b . co m / p a p e rs / m a n c a s - robust reading competition: Challenge
0198 natural.pdf / Natural Scene Text 1: Reading text in born-digital images,”
[11] XiaohangRen, Kai Chen, Jun Sun Understanding in Proc. IEEE Int. Conf. Doc. Anal.
Institute of Image Communication and [25] A survey of modern optical character Recognit., 2011, pp. 1491–1496.
Network Engineering, Shanghai Jiao recognition techniques Eugene [38] P. Shivakumara, A. Dutta, U. Pal, and C.
www.csi-india.org
24
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 R esearch  f ront

L. Tan, “A new method for handwritten for scene text recognition,”in Proc. Comput. Vis. Pattern Recognit., 2014,
scene text detection in video,” in Proc. IEEE Int. Conf. Comput. Vis. Pattern pp. 4042–4049.
IEEE Int. Conf. Front. Handwritten Recognit., 2014, pp. 4050–4057. [41] 41 J. L. Feild, “Improving text
Recognit., 2010, pp. 387–392. [40] 40 C. Yao, X. Bai, B. Shi, and W. Liu, recognition in images of natural
[39] 39 C. Leey, A. Bhardwaj, W. Di, V. “Strokelets: A learned multiscale scenes,” Ph.D. dissertation, Computer
Jagadeesh, and R. Piramuthu, “Region- representation for scene text Science, Univ. Massachusetts Amherst,
based discriminative feature pooling recognition,” in Proc. IEEE Int. Conf. Amherst, MA, USA, 2014.
n

About the Authors

Dr. Sharath Kumar Y H received the M.Tech and Ph.D. degrees from University of Mysore, Mysore, India in 2009 and 2014,
respectively. In 2013, he joined the Maharaja Institute of Technology, India, as a Assistant Professor. Now he is working
as Associate Professor and HOD, Department of Information Science. He has published papers in referred national and
international journals. His research interests include Data Mining, Machine Learning and Image Processing.

Mr. Puneeth P received the M.Tech from VTU, Belgavi, India in 2014. In 2015, he joined the Maharaja Institute of Technology,
India, as a Assistant Professor. His research interests include Data Mining, Machine Learning and Image Processing.

CSI Nihilent eGovernance Awards 2017

INVITATION

CSI SIG eGovernance team is inviting all the registered delegates,

Life Time Achievement Awardees, CSI Fellows, Executive Committee,
Conveners of all SIG’s, Chapter Chairs, Student Branch Chairs and
other invitees attending the

52nd CSI Annual Convention in Kolkata to grace the

15th CSI Nihilent eGovernance Awards 2017
distribution ceremony Scheduled from 1600 to 1800 hrs
on Saturday, 20 th January 2018

25
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 A R T I C L E

Make friendship with your Machine :

The HUMachine
Anila M.
Asst. Professor, MLRIT, Hyderabad

it, and getting what we need and want

through it.
It’s an out-and-out game-changer.
The swing of artificial intelligence
technologies—natural language
processing, computer vision, machine
learning, intelligent automation, robotic
process automation, deep learning and
more—will amend our relationship with
technology.
This changing relationship will
have major implications for how
businesses build systems. And it will
also unlock many new opportunities
to help professionals accomplish their
Artificial intelligence (AI) is about on an Amazon Echo speaker at home. personal and professional goals in their
autonomous learning and mimics the In each of these examples, people are lives and at work.
way a human brain learns: looking interacting with technology through an Fade your fear from automation:
for patterns within large amounts AI intermediator. And so, It’s a curator! Few examples which led the process
of data, such as speech, images, That helps you make your own machine automated
text, or anything else. And, just as a friend. Following figure tells you how There are many examples in
neuroscientists do not yet know exactly the AI works in machines. which machines automate the process
how the brain works, data scientists First it takes role of curator i.e. that humans do, like simplify hiring
also often struggle to explain how AI suggests what can be done basing on processes, from sorting resumés to
works. Technology simply moving too previous data, then will act as an advisor scheduling interviews
fast to be complacent. which guides users to make steps This technology-led process casts
AI will quickly and undoubtedly towards best results and later changes a wider net for talent and covers more
become more than an underlying its role to an orchestrator, where volume in lesser time than humans
technology capability, will pervade consumes past behaviour and integrate can. Besides, it targets not just active
intelligent enterprises and advance to a different tasks in-order to yield best jobseekers but even passive talent—
fundamental tool for daily engagement outputs. those who haven’t even thought of
with people. This is an intuitive way to relate quitting their jobs yet. And connecting
Because AI technology has already to technology. In fact, we’re already with them is merely a matter of clicking
marked its wider role in user experiences started treating technology like we a button.
across different interfaces. Think of treat each other(human). So instead of It’s not that workers have nothing
people’s conversations with chat bots on browsing a website to find information to fear from automation, but rather that
the web, their mobile phone queries to or a mobile app to finish the tasks, we’re companies will have a fair amount of
Google Now, and their requests to Alexa talking to the AI, making decisions with choice over what they want to do with
the extra efficiencies that technology
will bring.
Advisor Orchestrator “Journalists, beware!” is  a story
Curator
Learning from but Learning from past about a Google-funded software project
Suggesting relevant that aims to automate local news
also taking action action and collaborating
options based articles. Similar warnings have been
or guiding the user tasks across multiple
on previous user given to other knowledge workers:
towards an optimal channels to achieve
behavior. Investment bankers aren’t safe.
outcome. desired outcomes.
Auditors aren’t safe. Insurance brokers
aren’t safe. Therapists aren’t even safe.
www.csi-india.org
26
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 A R T I C L E

Nobody’s job is safe from the machines. understand what you want, the better years, everything quantifiable especially
This was supposed to be the point the world will end up being.” in finance and investment will be the
of Hammond’s computer science for And, on a positive side, “Something robots’ domain. But at the same time,
journalist’s course. “You have to use you’re in partnership with doesn’t trust will become the ultimate currency.
technology to do what you want to replace you.” I hope it’s right. The rise of AI robots in our industry
do,”. “The more you know how to use The future abstractions and models will — ironically and necessarily — re-
the technologies and the more you are far beyond our comprehension. In 20 humanize what we do.
n

About the Author

Ms. Anila M is currently working as Asst. Professor at MLRIT, Hyderabad and research scholar at KL University, Guntur. She
have 8.5 years of Teaching experience and her main interests are Databases, Data Mining, Artificial Intelligence, Machine
learning, and software Testing.

(ADVERTISING TARIFF)
Rates effective from April, 2014

CSI Communications
COLOUR
Colour Artwork (Soft copy format) or positives are MECHANICAL DATA
required for colour advertisement
Back Cover ` 50,000/- Full page with Bleed 28.6 cms x 22.1 cms
Inside Covers ` 40,000/- Full Page 24.5 cms x 18.5 cms
Full Page ` 35,000/- Double Spread with Bleed 28.6 cms x 43.6 cms
Double Spread ` 65,000/- Double Spread 24.5 cms x 40 cms
Centre Spread ` 70,000/-
(Additional 10% for bleed advertisement)
Special Incentive to any Individual/Organisation for getting sponsorship 15% of the advertisement value.
Special Discount for any confirmed advertisement for 6 months 10%.
Special Discount for any confirmed advertisement for 12 months 15%.
All incentive payments will be made by cheque within 30 days of receipt of payment for advertisement.
All advertisements are subject to acceptance by the editorial team.
Material in the form of Artwork or Positive should reach latest by 20th of the month for insertion in the following month.
All bookings should be addressed to :

Computer Society of India

TM

Unit No. 3, 4th Floor, Samruddhi Venture Park, MIDC, Andheri (E), Mumbai-400 093.
Tel. 91-22-2926 1700 • Fax: 91-22-2830 2133 | Email: [email protected]

Benefits for CSI members: Knowledge sharing and Networking

ƒƒ Participating in the International, National, Regional chapter events of CSI at discounted rates
ƒƒ Contributing in Chapter activities
ƒƒ Offering workshops/trainings in collaboration with CSI
ƒƒ Joining Special Interest Groups (SIG) for research, promotion and dissemination activities for selected domains,
both established and emerging
ƒƒ Delivering Guest lecturers in educational institutes associated with CSI
ƒƒ Voting in CSI elections
ƒƒ Becoming part of CSI management committee

27
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 A R T I C L E

Imaging Applications using

Machine Intelligence
Akila K. Jayasri P.
Asst. Professor of CS and Engg. at R.M.K College of Student of CS and Engg. at R.M.K College of
Engg. and Technology, Chennai Engg. and Technology, Chennai

In the recent times, developments in image processing have made enormous availability of information
in multimodal format. Although existence of sophisticated image processing tools analyzes the data,
these tools are complex and require intelligence to operate on the data effectively. For extraction of
useful information from the data, a user must possess appropriate intelligent machines and image
processing techniques. This article describes the use of machine intelligence to showcase various
image processing applications for data discovery from images.

Introduction
Machine Intelligence process development, creation of new agriculture, disaster prediction, and
explores application and theory oriented markets and increase in economic weather prediction. Intelligent image
knowledge which helps in language and growth. evolves scanning and multichannel
speech translations, visual processing, input, progressing with content analytics
ranking and prediction. Artificial Intelligent Image processing: and other linguistic enhancements.
Intelligence uses Machine Intelligence Intelligent Image Processing is For achieving the benefits of image
to solve most of the challenges faced in the fifth generation of industrial vision processing, we require a group of
the society. Machine Intelligent systems systems which possess the ability to have various technologies like scan and
are capable of performing intelligent a shared communication experience capture images, business process
tasks but using these systems we can among users using devices built inside management capabilities, repository for
perform operations at a higher level spectacle frames. This process is managing the data. Intelligent imaging
due to the speed and power of modern harmless to the human eyes. Intelligent provides a platform for connecting
computers. The major precursors of imaging is required for military and customer inputs with business system
Machine Intelligence are goal seeking, various types of surveillance, statistical leading to improvement in response
homeostasis, learning abilities. The data collection in the fields of forestry, time and anticipation of what customer
initial requirements of machine • deep learning
intelligence were state oriented, value Machine • unsupervised
learning • Supervised
driven, perpetual learning and self-
organization. The various usefulness Natural • text generation
obtained from machine intelligence are: Language   • question
ƒƒ Error reduction in terms of tasks Processing answering
performed.
•Image
ƒƒ Daily applications can be performed Vision recognition
efficiently • machine vision
ƒƒ Repetitive jobs can be handled
artificial • Adaptive
making it less time consuming intelligence Expert learning
ƒƒ Medical operations can be easily System • Reasoning with
executed. uncertainity
The various domains of artificial • Voice
intelligence are described below: recognition
Speech • language
In the near future, Machine translation
Intelligence will guarantee: Planning
ƒƒ Home advanced machines-robots, &
intelligent gadgets Scheduling
ƒƒ Machines having higher than Robotics • Automation
human intelligence providing
drastic progress in science
Fig.1 : Artificial Intelligence domain
ƒƒ Initiation of technology
www.csi-india.org
28
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 A R T I C L E

needs. Some of the benefits of image recommending trending filters and Google Cloud Vision API:
processing include : hash tags. ƒƒ Users can upload their images to
ƒƒ Faster access to information ƒƒ Cornea also contains basic filtering the Cloud Vision API to understand
ƒƒ Accuracy in data captured tools for users to be used on the contents of those images.
ƒƒ Fast imaging process their photos containing unlimited ƒƒ This helps in detecting objects as
ƒƒ Sustainable outcomes effects. well as to reading of text within the
ƒƒ Huge volume of data can be ƒƒ With Cornea, the ultimate aim is to image or identifying product logos.
scanned provide an accessible usage for all, ƒƒ With the beta release of Cloud
Intelligent Image processing works so that sharing a photo is simple. Vision API, one can access the API
on the principle of EyeTap principle. with location of images stored in
Google Cloud Storage along with
Bone Conduction
Headphones
Specialized low-
power processor existing support of embedding an
chip for computer
vision and deep image as part of the API request.
Battery &
Storage
learning
ƒƒ Additional capabilities to identify
the dominant color and broad set
of objects are exist in the image.
- 2D Camera (Prototype)
- Advanced 3D depth
sensing Camera (Planned)
Major features of this API include :
ƒƒ Inappropriate content detection
ƒƒ Identification of emotional
Fig. 2 : Eyeglasses attributes of people
ƒƒ Detection of text within images
It acts as an eye to record the Fig.3 : Cornea AI Application
scene available to the eye as well as a PhotoFy
display. It captures the world around it Face2Gene: ƒƒ It is a social photo editing and
and augments the image allowing it to ƒƒ It is a chamber of phenotype branding application editing about
overlay computer generated data over applications that provide 1,50,000 photos a day created by a
the top of normal view the user sees. comprehensive and accurate wide range of users.
It replaces each ray of light passing genetic evaluations. ƒƒ PhotoFy indicates potentially
through the lens of at least one eye of ƒƒ The patient’s facial description violent and adult content on user
the wearer, with a synthetic ray of light is compared to syndrome gestalt photos along with their abuse
responsive to the output of a processor. (gestalt scores) to quantify policies and is one of the most
It enables a shared visual experience similarity resulting in a prioritized powerful photo content creations
that connects multiple perceptions list of syndromes with similar in the world.
of individuals into a collective analysis.
consciousness. ƒƒ By combining the gestalt scores
WearTEL Technology: and the glossy features we result
ƒƒ It is a handled device that allows in clinically-important variants for
tapping of our eye and turning it better elucidation.
into a camera making a single ƒƒ It contains diagnosis and analysis
person perspective of creating of the syndrome, along with editing
personal documents. features, heat map, gestalt meter
ƒƒ The WearTel phone uses EyeTap indicating low, high and medium as
technology to allow individuals to parameters.
see each other’s point of view.
ƒƒ It allows each other to see what Fig. 5 : PhotoFy application
they are looking at other than only
the view of the person. Image Insights:
Cornea AI application: Brandwatch company has
ƒƒ Machine Intelligence has enhanced developed an application for business
the possibilities of automatic photo organizations developing products
editing / enhancement through to detect their product’s logo shared
Cornea application. online in the form of images. Image
ƒƒ This application makes use Insights is recognized as the world’s
of neural networks to find the leading logo detection technology. On
popularity score of photos before a daily basis, approximately 800-900
share them on social site and images are uploaded that contain logos
helps to improve the score by Fig.4 : Face2Gene Application of companies. for every company, there
exists a dashboard that is updated in

29
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 A R T I C L E

real time, contains information such harmful content and drawbacks. In the near future, the
as the location, at what time and how ƒƒ Find the significant people satellite imaging and remote sensing
the logo is shared. Logos are identifies promoting the brand application programs will feature a
even if it’s partially obscured, stretched variety of sensors orbiting the earth.
or edited. Using this information, a Conclusion: In the field of digital communication
company can: New upward trends in the area of and telecommunication, major uplift
ƒƒ Evaluate the performance of image processing prevails which are has significant relevance to image
advertisements used to perform wide range of knowledge processing. Machine intelligence will
ƒƒ Identify user-generated content discovery. Adoption of machine continue to change the world of creativity
ƒƒ View how customers use the intelligence techniques in image engaging learning accompanied by skill
product processing would develop only with the development.
ƒƒ Discover misuse of logo or any public opinion regarding the benefits
References:
[1] Cornea Application- https://cornea.ai/
[2] Face2Gene- https://suite.face2gene.
com/
[3] Intelligent imaging applications-
https://www.intelligent-imaging.com/
methods-applications
[4] Ram Ramanatha, Product Manager,
Google Cloud Platform[February 18
,2016]. Google Cloud Vision API
[5] Steve Mann, University of Toronto[April
23-2002]. EyeTap technology and
WearTel device in “ Intelligent Image
Processing”
[6] Image Insights- https://www.
brandwatch.com/image-insights/
[7] Embodied Intelligence- PowerPoint
presentation by Satpreet Arora
(07d05003), Rachit Gupta(07d05008),
Devendra Shelar(07d05010), Amrose
Birani(07005003), CSE, IIT Bombay.
[8] Intelligent Imaging for improved
banking performance and profitability-
Kevin Crane, MBA
Fig. 5 : Image Insights for Nike n

About the Authors

Ms. Akila K is the Assistant Professor of Computer Science and Engineering at R.M.K College of Engineering
and Technology, Chennai. She completed her B.E (CSE) in the year 2003 and M.E (CSE) in the year 2005 from
Sathyabama University. She is pursuing her Ph.D.(CSE) in Anna university. Her areas of interests are Data
Mining, Image Processing, and Pattern Recognition. Her area of research is Digital Image Processing.

Ms. Jayasri P is the student of Computer Science and Engineering at R.M.K College of Engineering and
Technology, Chennai. Her areas of interest include Web Services, doing Mini Projects, participating in various
inter college activities.

Like CSI on facebook at :

https://www.facebook.com/CSIHQ

www.csi-india.org
30
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 A R T I C L E

Understanding Machine Vision by

Human Activity Recognition
Dushyant Kumar Singh
Asst. Professor, Department of Computer Science and Engineering, Motilal Nehru National Institute of Technology Allahabad, Allahabad, UP

Introduction
Video data is present in insanely technologies [3]. The consequence an additional challenge to the
huge amount in every part of the world is that, instead of employing video recognition model. For example, the
today. A huge volume of video data is surveillance mainly as an after-effect motion during the transition period
produced each and every second by forensic tool, it is now feasible to deploy between two activities is difficult
surveillance cameras, video cameras, digital, network-based surveillance for the underlying classification
mobile devices, etc. Public surveillance systems to provide interactive, real- algorithm to recognize.
cameras are the single biggest time monitoring and surveillance. 3. Insufficient Training Set: collection
source of all the videos in this world. An effort is made here to propose a of training data under realistic
This presents us with a challenge of computer framework for video analysis conditions is very difficult. As
maintaining and categorizing all of that to enable robust and real-time human mentioned in the subject sensitivity
data efficiently and in a human-friendly activity recognition. Literature provides challenge part, it is highly desirable
manner for easy storage and retrieval. some existing proposals on techniques that the training data must contain
Apart from this, a system capable of for human activity recognition [1, 2 & 3]. as many varieties of the subjects as
recognizing activities from a video Challenges possible. However, it is not easy to
can enable us to develop a partially or There are still many issues coordinate people of different ages
fully automated surveillance system. and challenges that motivate the and body shapes to collect data
The underlying principle remains the development of new activity recognition under a controlled lab environment,
same for every application, i.e. tagging techniques to improve the accuracy not to mention the varieties of the
each second of video with a label that under more realistic conditions. environment itself.
represents the activity happening in it 1. Subject sensitivity: The accuracy
at that time instance. A system capable 4. 
Personal Biometric Signature: A
of activity recognition, especially
of recognizing the human behavior subject motion pattern is usually
those based on the accelerometer
can have several applications, from exclusive and unique. For example,
data, is heavily affected by the
visual surveillance to automatic sports when people raise their hands, it is
subjects participated in training
commentary. In particular, a method for almost impossible for two people
and testing stages. This is mainly
determining an instantaneous human hands to share the exact same
due to the fact that different people
action that may comprise several motion patterns.
have different motion patterns.
actions in sequence would inevitably be 5. Computational Power: Neural
a core building block of the system. Efros 2. Activity Complexity: The complexity
Networks require a great deal of
et. al. [4] has proposed some technique of user activities also brings
for recognizing person-centred actions INPUT VIDEO FRAGMENT INTO SERIES
like walking, running, and standing. DATA OF JPEG IMAGES
Such a system can also help in
content based storage and retrieval of
video data in any vision based system.
The core functions of such system GENERATE SEQUENCES USING
could be, recognizing various human- SET OF IMAGES BY PASSING
object, human-human and object- THEM THROUGH A CNN
object interactions etc. Based on these
findings any video can be tagged with
appropriate labels. These systems can
be deployed in security applications
for surveillance. Since the introduction USE THE LSTM-RNN TO
of analog video surveillance systems
BUILD AND TRAIN RECOGNIZE ACTIVITIES FROM
LSTM-RNN UNSEEN VIDEO DATA
back in the 1970s, tremendous strides
have been made in sensing, storage, Figure 1: Human activity classification in a video using LSTM-RNN
networking, and communication

31
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 A R T I C L E

computational power to run. Most higher training and classification time

of the big neural networks run on was recorded on CPU based machines.
giant network of GPUs with several While this time was reduced by using
Gigabytes of main memory. GPU based NVIDIA Jetson TK1 SoC
board.
Methodology
The use of machine learning References
(a) Biking [1] Caba Heilbron, Fabian, et al.
techniques could be one solution domain
“Activitynet: A large-scale video
to cover the above stated challenges benchmark for human activity
for human activity recognition. To understanding.”  Proceedings of the
handle activity complexity and learning IEEE Conference on Computer Vision and
inputs in terms of huge dataset, the Pattern Recognition. 2015.
Convolutional Neural Network (CNN) [2] Bao, Ling, and Stephen Intille.
[5] and Recurrent Neural Network “Activity recognition from user-
(b) Fighting (RNN) [6] are the tools. Long Short- annotated acceleration data.” Pervasive
Term Memory (LSTM) [7], a special kind computing (2004): 1-17.
of RNN is used in a technique proposed [3] Kim, Eunju, Sumi Helal, and Diane
by my students for human activity Cook. “Human activity recognition
classification. The flow graph in figure 1 and pattern discovery.”  IEEE Pervasive
shows the procedure of human activity Computing9.1 (2010).
(c) Hand Gesture classification in a video using LSTM-
[4] Efros, Alexei A., et al. “Recognizing
RNN. action at a distance.” null. IEEE, 2003.
The experiment was done for
classifying the video clips based on [5] Yang, Jianbo, et al. “Deep Convolutional
Neural Networks on Multichannel
the activities it posses. The classes
Time Series for Human Activity
for which LSTM-RNN was trained Recognition.” IJCAI. 2015.
were Biking, Fighting, Hand Gesture,
(d) Jumping Jumping and Running. Same are also [6] Gregor, Karol, et al. “DRAW: A
recurrent neural network for
shown in figure 2, when applied on
image generation.”  arXiv preprint
sample video clips. arXiv:1502.04623 (2015).
Conclusion [7] Sak, Haşim, Andrew Senior, and
LSTM model is designed for Françoise Beaufays. “Long short-
recognizing 5 different activities. This term memory recurrent neural
model makes best use of sequential network architectures for large scale
(e) Running nature of video data. The accuracy acoustic modeling.”  Fifteenth Annual
Fig. 2: Example of predictions of Activities achieved is high, which accounts to an Conference of the International Speech
using CNN models DataSet http://crcv.ucf. average of 95%, with our considered Communication Association. 2014.
edu/data/UCF101.php
dataset and the set of classes. The n

About the Author

Dr. Dushyant Kumar Singh (Membership No. I1500373) Areas of my research interest are Computer Vision,
Image Processing, Embedded Design, Advance architectures etc. I have good number of research publications
in reputed journals and conference proceedings. I frequently write articles for different magazines. Besides
authoring papers, I am also the reviewer for good flagship conferences and reputed journals.

www.csi-india.org
32
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 A R T I C L E

Role of Machine Intelligence for

sustainable agricultural development
Mahendra Balkrishna Salunke Ashish V. Salunkhe
Dept. of Computer Engg., Pimpri Chinchwad College of Engineering Dept. of Computer Engg., Pimpri Chinchwad College of Engineering
& Research, Ravet, Pune. 412101 • E-Mail: [email protected] & Research, Ravet, Pune. 412101 • E-Mail: [email protected]

This article expounds an overview of Machine Intelligence and its broad spectrum over different fields
and aspects in near future and it emphasizes its significance in the field of agriculture, its impact
in optimizing the farming practices. It highlights the associations, organizations working in Machine
Intelligence to achieve a sustainable autonomous farming strategy.

Introduction
Homo sapiens—man the wise. intelligence and relation between them 90-95%. At that very phase, resource
Right from the beginning of human were more extensive and sophisticated. constraints would arise to an alarming
race, human being has consistently They are seen to be well-considered level.
tried to decode how he thinks, how a and sound in appraisal. So, looking at all the states and
mere handful of matter can perceive, So today Machine Intelligence also knowing the chances, giant strides are
comprehend, predict and maneuver an referred to as Artificial Intelligence has put to seize maximum profit and yield
environment that is more outsized and a layout as with minimum use of resources with
complicated than itself. The impact of AI and its cognitive the help of innovative technologies that
The present scenario where AI consequences over years, has not enable cultivators a broader yield.
(Artificial Intelligence), Automation has only allowed to cut down manual So how would machine intelligence
a upper hand over traditional set up, calculations significantly, but through improve agricultural aspects?
machine intelligence lays emphasis on automation, gradually and precisely
building smarter, resilient and efficient predict results for future outcomes. Farm Analytics
entities and not on only studying them. So, while technology was meant for So what is farm analytics?
So backtracking to 1950, Alan cognitive applications in industries, the As agriculture involves parameters
Turing, a British mathematician, came progressive onset of global warming and like weather, soil type, water resources,
up with Turing test which set up a climate change, over the last decade, time; analytics begins with arbitration
platform for abilities of a machine to has gradually allowed AI to lance into of all these parameters and later on
think, in his paper “Computing Machinery a traditional sector like agriculture, to using all these factors for prediction of
and Intelligence”. help cope with the intricacies in modern crop production process, determining
Alan Turing said, “We can only see day farms. valuable insights, developing a broad
a short distance ahead, but we can see Agriculture is the sector that dataset, generating insights that
plenty there that needs to be done. supplemented the evolution of human are actionable, delivering them and
I propose to consider the question, race from centuriestill date and updating them on a regular basis.
‘Can machines think? thereby accomplished one of its most Farm Analytics with the help of
A computer would deserve to be basic needs: food. This still remains neural networks can enclose large
called intelligent if it could deceive a a fore motif, but it is cohesive since datasets, integrate various sources of
human into believing that it was human.” it is compelled by many factors that information and increase crop yield
Intelligence is a spectrum of include economic, environmental and and optimize resource usage, thereby
phenomena. Hence through recurrent humanitarian aspects. So agriculture cutting down the cost. Modern-Day
observations of evolution of intelligence, together is a $5 trillion sector, which technologies such as the Internet
it can be concluded that considering has 10% of global consumer spending, of Things and Cloud Computing
human intelligence, there have been 40% employment and near about can influence this development and
theoretical limitations, unlike the 30% of greenhouse-gas emissions. introduce machine intelligence in
evolution of machine intelligence. So Considering the present scenario, it is farming.  This is incorporated by using
in near future machine intelligence is likely that the in near future or over a Big Data, having huge storage spaces
likely to outpace human intelligence. span of 25 years the crop demand and that can be netted, analyzed and used
So coming back to Turing’s reflections its consumption byhumans and animal for decision-making. Innovations like
on machine intelligence, human feed would undoubtedly increase by the  Watson IoT platform  augment

33
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 A R T I C L E

value by introducing machine learning can interact with farmers. Thus System receivers, Yield monitoring
features to sensor data, converting Chatbots are conversational virtual and mapping, Grid soil sampling
management systems in real machine assistants who automate interactions and variable-rate fertilizer (VRT),
intelligence systems. with end users. AI powered chatbots, Remote sensing, Crop scouting,
Image Recognition using Neuro-Linguistic Programming Geographic information systems
Image recognition in Agriculture and machine learning techniques, (GIS) and Identification of a precision
includes fields such as fruit grading, understand natural language and agriculture service provider. Fields have
weed detection, poultry inspection, interact with users in an adapted way. variable yields across the landscape.
detection of fruit diseases, defects, and Though it is early for implementation it These variations are outlined to
contamination, spotting rodents and can prove to be another valuable asset. regulate practices on soil properties,
rats near granary and other outspread Mixed Reality Fields/Greenhouses environmental physiognomies. Such
applications. Mixed Reality, a technology, with physiognomies may affect texture,
Image Recognition or machine union of two virtual environments where assembly, moisture, nutrient status and
vision has the prospective to add two worlds coexist together. Basically it weather, weeds, insects and disease
valuable assets to the agricultural can be a combination of augmented and and landscape position.
industry. Image recognition can prove its virtual reality techniques. Mixed Reality Summary
value in expertise like remote sensing allows 3-D Mapping of actual field intoa Although early in days, this
for natural resources valuations, virtual environment where farmers can new foresight for farming can be
process automation, product quality, generate different setups or scenarios revolutionary and game changing
grouping and precision farming. of crop farming. Headgears, VR-AR one. Farmers with full access to these
Thermal imaging whose infrared headsets can turn 3-D mapping into sort of revolution can come up with
range lies between 3 to 14 μm detects a virtual environment and augment higher yields and profits and thereby
water presence. Water affects the physical objects. For example in increase economy and exports. AI,
thermal properties of plant. Water greenhouse system,Sensors, cameras Machine Intelligence, IoT, Big Data,
contents can be analyzed in pre etc. are mounted on PVC pipes located Cloud Computing, AR VR, and many
harvesting operations. Thermal imaging around the perimeter of the greenhouse. other forms thus can be revolutionary
also deals with recognition of nursery, All data is sent to the server and thus in every aspect though there might be
scheduling of irrigation, forecasting of machine learning algorithm routes some clauses that can arise but they
yield, level of greenhouse gases, attacks this information and evaluates the surely won’t be at an alarming level.
of rodents, termites. Post harvesting whole process. Thus Mixed Reality Organizations and companies like PEAT,
operations include assessment of uses AR-glasses or VR headsets that Trace Genomics, Gamaya, aWhere,
maturity, discoloration detection, allow farmers to see the plant with plantix and many others have come up
detection of extraneous substances in augmented objects. with smart agriculture concepts and
vegetables, fruits. Thermal imaging Precision Agriculture technologies. Investors invested more
cannot be that helpful as climatic Precision agriculture is an than $3.2 billion in agriculture in 2016,
conditions and plant physiology varies integrated crop management system $200 million by Plenty and $305 by Blue
depending on landform. that tries to replicate the amount of River, both agtech companies, showing
Chatbots inputs with the actual crop needs for the attention and investment drawn by
Virtual assistants are readily trivial areas in a farm field. this sector. Thus agriculture sector can
available in smartphones and they So involves Global Positioning grow in a broader aspect. n

About the Authors

Mahendra Balkrishna Salunke (CSI Membership No: I1504453) is an Assistant Professor & Head, Department
of Computer Engineering, Pimpri Chinchwad College of Engineering & Research Ravet, Savitribai Phule Pune
University. He received Master of Technology in Computer Science & Engineering from PDEA COE Gulbarga,
VTU, Begaum. He is having total experience of more than 18 years. He is a member of ACM, Life Member
of CSI, Life Member of ISTE. He is having research interest in Embedded Security, Internet of Things, Deep
Learning.
Ashish V. Salunkhe is pursuing Bachelor of Computer Engineering at Pimpri Chinchwad College of Engineering
& Research, Ravet, Pune, Maharashtra. He is a Student Member of ACM and Students Coordinator of PCCOE&R
ACM Student Chapter. His area of interest and specialization is in Machine Learning and Image Processing.

www.csi-india.org
34
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 A R T I C L E

Machine Intelligence Challenges in

Military Robotic Control
S Balakrishnan D Deva
Associate Professor, Dept. of IT, Sri Krishna College of Engineering Third Year, Dept. of Mechatronics Engg.,
and Technology, Coimbatore, Tamilnadu, India. Sri Krishna College of Engg. and Technology, Coimbatore, Tamilnadu, India.

The Youngest Innovation of Universe

OUTPUTS
The youngest innovation in the INPUTS F (X)
world of technology is the intelligence
that is incorporated in machines, that
fulfils many industrial and smart home
requirements. Machine intelligence LEARNING
may involve in two aspects, the artificial DATA TRAINED MACHINE
intelligence and the machine learning.
ALGORITHM
It was a great evolution in the field of
defense, when machine intelligence
found to be applicable in its various
activities.
Why Machine Intelligence in Military NEW DATA
Robots?
The development of science and
technology is abruptly spreading in all
the sectors nevertheless the defense Limitations due to Human Interface by altering the machine intelligence
force failed to use the new innovations Besides its usefulness, every new embedded in to it may cause
around the globe. The backbone of a innovation has its own limitations. unimaginable loss to the military
nation is determined by its defense Some of the limitations are neglected armies. When the original intelligence
strength, but the personal losses to the but some give rise to serious issues programmed to the machine is changed
military people are inevitable. In order in the society. During the time of war, by the victims, there are the following
to meet these losses every military when the military robots are attacked challenges: The autonomous robots
force decided to use the autonomous by the opposite army, at the same place can kill its own army persons by its
robots which perform all the actions they have the ability to rehabilitee at wrong recognition. The robots get
and duty that a trained military man the same place and continue its work, confused with right from wrong and it
does, but the percentage of demerit thus they no longer have the control may not attack the persons, whomever
is more when compared to the merits of humans and become independent it considers as its own army and it could
of using autonomous robots. The robots. Though this machine intelligence be a great advantage to the opposition
rise of demerit is due to the lack of is invited in the field of robotics, the army. One of the major risks in hacking
“intelligence” that the robot (“machine”) scientists and defense officers are the machine intelligence is that, the
has. The incorporation of such useful taking a step backward to employ these victim may not only turn our own robot
intelligence into the robot paves way autonomous robots in the battle field against us, but also he can make it
for it to perform more efficiently and due to its lack of human interface. The explode, which is a great loss at times
some times more than a military man. put forth the report that during the war, of war.
With the help of machine intelligence these robots cannot cooperate human Causes of Machine Intelligence
it decides whether to fight or flight at army due to the human interface issue Disadvantage
times of risky operations. and will cause unity problem in the Normally when a code in embedded
force, which is to be strictly avoided at systems or in any programming
times of war. Other than military, this language is written, it performs only that
possibility asks many questions in the particular task, thus it does not bother
field of automation, whether to trust the about the external tasks, whereas the
machine intelligence whole heartedly, machine intelligence takes the sole
even though they are no more humans. responsibility to the unexpected tasks
Hacking Issues in Machine and it volunteers itself for the execution
Intelligence of the same. At what so ever be the
The hacking of military robots circumstance, the machine intelligence

35
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 A R T I C L E

in the autonomous robots always failed the below mentioned qualities to be information’s must be updated in the
to win the human intelligence. more efficient. regular basis to improve the intelligence
Emotional Intelligence to Machine of the machine, pessimistically this
MOTIVATION
Intelligence kind of updating costs very high for the
Many researchers around the world armies to spend on the machines.
are more bothered about the emotional Even money can be compensated
SOCIAL by the army, but not the time! Yes, the
EMATHY
quotient of machine intelligence added SKILLS
with the general intelligence that is ROBOT machine intelligence needs more time
seen in every machine. Though machine for any up gradation and in case of
intelligence is more than enough for a servicing the software, it consumes
machine to do wonders, it must be more time, which is not entertained in
emotionally satisfied too, since it is SELF SELF the battle field.
AWARENESS REGULATION
going to work along the humans who Conclusion
have emotional intelligence. As far as The great and rapid evolution in
the percentage of emotional intelligence Economic and Time factors the field of automation by the machine
is taken in to account in military robots, Fabrication of a simple robot itself intelligence is inevitable in the world of
it is very minimal or sometimes nil. This is economically very high and in case technology, nevertheless it should not
reason may stand for an excellent proof of a autonomous military robot with compensate the security of the people
to limit the usage of autonomous robots the machine intelligence would be and the machine-to-man relationship,
in the army. Even though emotional more higher. This paves no way for the since each and every development in
intelligence incorporation is not that economically poor countries to have the universe is only for the betterment
much challenging to the innovators, their own military robot in their army. of human life. As far as machine
on the other hand allowing that such The reason behind the high cost is the intelligence in robots are concerned,
intelligence to a machine that can software used and the programming proper allotments of the autonomous
be hacked, remains the reason for done to think intelligently. robots with respect to the place of
non-incorporation of the emotional Every human need to be updated to operation and the type of function must
intelligence to the machine intelligence. be smart enough, then the same is for be carefully monitored in order to use
The military robots must have all our military robots too! Each and every machine intelligence in a better way.

About the Authors

Dr. S Balakrishnan (CSI Membership No. 1187923) Professor, Department of Information Technology, Sri
Krishna College of Engineering & Technology, Coimbatore, Tamilnadu, India. He has 17 years of experience in
teaching, research and administration. He has published over 15 books, 3 Book Chapters, 3 Technical articles
in CSI Communications Magazine and over 30 publications in highly cited Journals. His professional awards
include: Teaching Excellence Award, I2OR - Bright Researcher Award, Best Outstanding Faculty Award, Best
Teacher Award, Best Research Paper Award, Best Book Publication Award and Best Book Chapter Award,
Special Contributor Award and Star Performer Award. His research interests are Artificial Intelligence, Cloud
Computing and IoT. He has delivered several guest lectures, seminars and chaired a session for various
conferences. He is serving as a Reviewer and Editorial Board Member of many reputed Journals and acted
as Session chair and Technical Program Committee member of National conference and International
Conferences at China and Bangkok. Dr.Balakrishnan is a life member of ISTE, IAENG, IEAE, IARDO, CSI,
UACEE, SDIWC and CSTA.
Mr. D Deva is pursuing Third year Mechatronics Engineering in Sri Krishna College of Engineering and
Technology, his area of interests are Thermodynamics and conservation of Energy, he has authored a
motivational book “The Ladder For Teens”, and also delivered various extramural presentations and Technical
seminars. He has published the articles in various International Journals.

www.csi-india.org
36
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 SECURIT Y CORNER

Cyber Crime and Social Security –

The Dark side of Internet
Priyanka Tomar
[email protected], [email protected]

This paper explains the dark side of the Internet and research on cyber crime cases along with the
current cyber crime challenges and its effect on social security. Cyber crime has taken organized shape
and fraudsters are committing it professionally. Many persons were duped of a total of nearly millions
by fraudsters. Cyber crime increased post-demonetization, after the Rs 500 and Rs 1000 note ban,
people switched to various online payment methods and commonly used method is mobile wallet. This
led to a lot of cyber criminals taking advantage of the opportunity and scamming people of their credit
card, debit card and net banking details. As per research, economic fraud tops the list of cybercrimes
in India. Techniques used by cyber criminals have been explained in this research paper e.g. how and
why cyber criminals gather personal information of the people, why Cyber Criminals hack into victim’s
Facebook account, why we receive third party weblinks into our emails and how are they being a big
threat to social security along with “ Do and Don’ts” to create awareness among the people. The dark
side of Internet is that cyber criminals not only steal the information to make online purchases but
they sell victim’s personal information along with his debit/credit card details over Deep web. One can
buy arms, drugs or hire/rent criminal services over deep web.

Cyber Crime and Social Security- The banking has also lead to the increased cameras. 2016 was popular for Ransom
Dark side of Internet cybercrimes because bank accounts ware Attacks.
Nowadays Cyber Criminals could be hacked and money could be Why Cyber Criminals gather Personal
work in a very organized manner and transferred into another account all Information-
normally are highly qualified. The together by finding vulnerabilities in They use this information to sell one
online community is evolving at fast the application or via phishing email country data to another, one corporate
pace and have created a parallel world or hackers may try to steal the back data to another business rival, commit
of Cyber Crime. Because of increased account details. Though banks are ID fraud to escape out of the country by
cyber crimes, the social security of upgrading their technology to make spoofing the actual ids, banking frauds,
people is in danger, cyber criminals transactions safer. It lead to many ways to lure common person for lucrative job
not only hack and transfer money from by which one in present time commits offer, to trap teenagers to commit crime
e-banking accounts, mobile wallets Cybercrime, hence, Cyber Crime affects for them.
illegally but they sell banking data social security. If cyber criminal has some of your
thereafter. Likewise hackers hack the With the growth of e-business, personal information like date of birth,
email ids or social networking websites cyber criminals not only exploit well birth place, best friend name, school
accounts  to take revenge of personal known existing vulnerabilities but name etc then it becomes very easy for
enmity . Now cyber crime has taken a also find new loopholes in the web them to get access of your email id. Let
form of Cyber Terrorism and almost application, online software and us understand it by an example:
on daily basis cyber war incidences hardware too. CCTV cameras are being X has an email id [email protected]
are also being reported by the media. used as a tool to hack a server, hackers and at the time of account creation on
Defaming someone, or cyber terrorism installs malware into CCTV, as they rediff, one has to answer some secret
or cyber war cases affects society in have running Linux server and hackers questions which are useful when
a very adverse manner. Due to the are able to execute scripts like jquery, someone forgets the password and
increased use of the social networking cross-site scripting etc. and cyber wants to retrieve it. Normally security
sites many accounts of the respective criminals get access in to government questions are Birth Place, City you are
users are under the scanner as and corporate network. born, Best Friend name, First School
advancement in technology has lead to One example is - Botnet Attack. name etc. And all this information is
innovative hacking methods. The online These can be performed using CCTV

37
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 SECURIT Y CORNER

available on social networking websites social networking accounts or information online, violence threat etc.
like Facebook. Cyber criminal can sometime a phone call can be fruitful. Cyber stalking of intimate partners is
gather all this information easily by the online harassment of a current or
Step 2 : Cyber criminals send an email
being befriended with X and later on can former spouse, boyfriend, girlfriend.
to that person and ask him or her to
get into X email id. Harassers send repeated insulting
make a payment against some genuine
Methods used by Cyber Criminals to looking invoice or reason. They use the or threatening e-mails to the victim,
gather the Information familiar language in the emails which monitor or disrupt their e-mail use,
Social Engineering very popular may give feel that some well known and use the victim’s account to send
method and cyber criminals are business vendor is communicating. e-mails to others posing as the victim
successful most of the time, they use or to purchase goods or services. Cyber
Step 4 : The cyber criminals receive stalkers use the internet to research and
various Search Engines - like Google,
bing, shodan etc, Social Networking the funds and disappear and at times compile victim’s personal information,
websites - like face book, twitter, without leaving the trail. to use in order to harass her or him.
LinkedIn. As per Verizone 2015 DBIR According to the Symantec 2014 With the advent of technology, cyber
Report humans, or users, account for Internet Security Threat Report, Social crime poses as a major threat to the
90% of security incidents: Engineering Fraud has sharply risen security of a person as a whole.
Google Dorking - Very powerful over the last few years. Cyber criminals Type of Cyber Crimes - Amongst
information gathering technique, have discovered how easily this scam the various cyber crimes committed
google dork is a search string that can succeed with minimal cost and against individuals and society are-
uses advanced search operators to find effort. Their study shows that Social 1.  e-mail hacking
information that is not readily available Engineering Fraud attacks: 2.  Cyber stalking
on a website. Google dorking, also ƒƒ Increased 91% between 2014-2015 3.  Cyber pornography
called as Google hacking, can return ƒƒ Occurred at a rate of over 100,000 4.  Defamation via Image Spoofing
information that is difficult to locate attacks each day 5.  Email Spoofing
through simple search queries. ƒƒ Target businesses of all sizes 7. Cyber Terrorism
Many people don’t secure their (34% of large businesses, 31% of 8. Phishing & Spam
online directories using strong password medium, businesses and 30% of 9. Social Networking Account Hacking
or at times they are open to access small businesses) Harassment includes blackmailing,
for everyone. They store their person ƒƒ Target 1 in 2 large businesses and 1 threatening, cyber bullying, and even
information like Family photographs, in 5 small businesses cheating via email. Cyber stalking
academics records, marks sheets, Few Examples of Cyber Crime involves following a person’s movements
passport copy, pan card etc. Such type Cases : These cases shows the across the Internet by posting messages
of open information is a big threat to strategy adopted by cyber criminals on the bulletin boards frequently visited
social security of that individual. and how availability of online personal by the victim, entering the chat-rooms
Stealing the Data in Transit - information can be a threat to social frequently used by the victim, constantly
People do not use proper encryption security. bombarding the victim with emails
methods so while data is in transit and it In a case, the Delhi high court etc. Cyber Stalking usually occurs
transits in clear text form then a hacker imposed a fine of ` 20,000 on a man with women, who are stalked by men,
can steal it easily using sniffers. for posting on a social networking site or with children who are stalked by
Tor - Tor is a free software enabling “objectionable” photographs and the adult predators or pedophiles. Cyber
anonymous communication. If cyber telephone number of a woman who had stalkers target and harass their victims
criminal uses Tor then it is very difficult studied along with him in the school. The via websites, chat rooms, discussion
to be traced back to him to know the court imposed the fine while quashing forums, open publishing websites and
kind of internet activity he has done. a criminal case registered against the emails.
Tor’s use is intended to protect the man under the Information Technology The recent being the case of
personal privacy of users, as well as Act (ITA).While the victim and the MK who was recently arrested by the
their freedom and ability to conduct accused may have reached an out-of- New Delhi Police. He was stalking an
confidential communication by keeping court settlement to avoid legal tangles Indian lady, R by illegally chatting on
their Internet activities from being and negative publicity, many cyber law the Web site using her name. He used
monitored but it being used by hackers experts say that a stringent punishment obscene and obnoxious language, and
too. should be awarded in cyber crime cases distributed her residence telephone
Social Engineering Email Fraud - since with widespread usage of social number, inviting people to chat with her
cyber-criminals are very good in it. networking sites there is a fear of cyber on the phone. As a result of which, R kept
Step 1 : Cyber criminals look for key harassment, and hard punishments can getting obscene calls from everywhere,
decision maker or authorized signatory discourage such type of cyber crimes. and people promptly talked dirty with
to within an organization. This can be Harassment and stalking of people her. In another case, an engineering
done either by gathering information is common, and include variety of and management graduate, facing
from their website or organization’s threats like posting of person’s personal prosecution in a dowry harassment
www.csi-india.org
38
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 SECURIT Y CORNER

case, was arrested by Delhi police for a complaint with the Kalkaji police 7. The duo would enter an
sending obscene e-mails in his wife’s station(South Delhi) after the duo used incomplete or fake address on the
name to several persons. Internet has his credit card to shop on Flipkart and e-commerce websites and get the
provided a medium for the facilitation Snapdeal. product delivered midway to avoid
of crimes like pornography. Cyber porn Investigations revealed that the suspicion.
as it is popularly called is widespread. duo had issued a duplicate SIM card of 8. They sold gold coins in open market
Almost 50% of the web sites exhibit the one owned by Gautam. On finding with the help of e-bills. They had
pornographic material on the Internet out about the duplicate SIM, the crime even sold coins to the ‘Cash for
today. The elementary problems, which branch swung into action. Gold in Delhi’ outlet in Delhi.
are associated with Cyber-Crimes, are The Duo Duped Citizens across the 9. Till now, five other victims have
Jurisdiction, Loss of evidence, Lack of Nation been identified and they have
cyber army and Cyber savvy judges who Police said the accused had been already lodged complaints.
are the need of the day. Judiciary plays siphoning off cash using credit cards 10. Police said that efforts are being
a vital role in shaping the enactment of several unsuspecting citizens across made to locate other victims with
according to the order of the day. Today the country for some time now. They the help of CDR analysis and data
with the growing arms of cyberspace duped citizens under the pretext of of e-shopping websites.
the growing arms of cyberspace the providing new credit card with enhanced 11. While Kushwah is a native of
territorial boundaries seems to vanish credit. Gwalior, Madhya Pradesh, Rawat
Cyber space is a transit space for The crime branch got a tip- is a native of Shivpuri in Madhya
many people, including offenders. While off about Kushwah’s presence near Pradesh.
people do not live in cyber space, they Videocon Towers in Jhandewalan. 12. Kushwah and Rawat have
come and go like any other place. This Acting on the information, they nabbed completed their Masters of
nature provides the offenders the chance Kushwah and later arrested Rawat from Computer Application.
to escape after the commission of cyber Motiyah Bag locality. Cops recovered 13. Police said that being a JAVA
crime. Human trafficking has increased two gold coins, nine mobile phones, 22 developer, Kushwah had made
along the border in WB, Bihar etc. SIM Cards, fake IDs including an Aadhar software programmes for as many
Cyber crime cops said that after card, four dongles and a computer from as 12 government hospitals in
December 12, cyber fraud cases have the accused. Delhi.
witnessed a steady increase. “Now we 1. The duo obtained details of people 14. Both Rawat and Kushwah were
are getting at least four cases of credit interested in getting a new credit classmates in their under graduate
or debit card theft every week. Apart card with enhanced credit limit course.
from these cases, we are also getting from the search service ‘Justdial’. FBI: $2.3 Billion Lost to CEO Email
One-Time Password (OTP) frauds as They registered with the online Scams - The U.S. Federal Bureau of
well,” said Raghuveer, ACP, cyber crime portal using the name ‘j2k system
Investigation (FBI) warned in April
cell. Source- Times of India pvt. Ltd’.
2016 about a “dramatic” increase in
As per ASSOCHAM, in a report in 2. They later visited the unsuspecting
so-called “CEO fraud,” e-mail scams
October this year, the number of mobile citizens and sought the details of
in which the attacker spoofs a message
frauds are expected to grow by 65% by their existing credit cards.
from the boss and tricks someone at
2017. Credit and debit card fraud cases 3. With the help of Delhi police’s
the organization into wiring funds to
have topped the charts of cyber crime online registration facility for NCR,
the fraudsters. The FBI estimates these
and increased six times during the last they would contact the customer
three years, they said. Source- Times of scams have cost organizations more
care of the mobile company used
India than $2.3 billion in losses over the past
by the victim.
3.2 million debit card details three years.
4. They would ask the companies
stolen as India faces one of its largest In alert posted to its site, the
to block the SIM cards to prevent
data breaches ever Source -Economic FBI said that since January 2015, the
the victims from getting any SMS
Times agency has seen a 270 percent increase
on their phones about online
Delhi Police busts credit card fraud, in identified victims and exposed losses
transactions.
arrests 2 techies- Delhi Police crime from CEO scams. The alert noted that
5. After blocking the original SIM,
branch busted an online credit card they would visit the outlet of law enforcement globally has received
fraud racket with the arrest of two people cellular company and with the help complaints from victims in every U.S.
on Monday. Jayshankar Kushwah and of a copy of the victim’s ID proof, state, and in at least 79 countries.
Rajesh Rawat, both residents of New obtained a duplicate SIM. What is CEO fraud attack
Delhi’s Sarai Rohilla with a Masters in 6. Before making any online CEO fraud usually begins with the
Computer Applications, were arrested transactions they verified the thieves either phishing an executive and
after the police received a tip off on available limit of credit cards. gaining access to that individual’s inbox,
their whereabouts. Based on the credit limit they used or emailing employees from a look-alike
A businessman Gautam was duped to purchase gold coins through domain name that is one or two letters
of ` 2.52 lakh by the duo. Gautam filed Amazon, Snapdeal, Flipkart, etc. off from the target company’s true

39
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 SECURIT Y CORNER

domain name. For example, if the target increase: (https://www.symantec.com/ most common thread vector, so be
company’s domain was “example.com” security-center/threat-report)- careful while checking your inbox-
the thieves might register “examp1e. “In 2015, we saw a record-setting A. Check “From” address
com” (substituting the letter “L” for the total of nine mega-breaches, and the carefully
numeral 1) or “example.co,” and send reported number of exposed identities B. Don’t reply to the email with the
messages from that domain. jumped to 429 million. But this number word “Urgent action required”,
Unlike traditional phishing scams, hides a bigger story. In 2015, more this may be malicious email.
spoofed emails used in CEO fraud companies chose not to reveal the C. Crosscheck the website
schemes  rarely  set off spam traps full extent of their data breaches. A and email address, in
because these are targeted phishing conservative estimate of unreported case of official email or
scams that are not mass e-mailed. Also, breaches pushes the number of records email representing some
the crooks behind them take the time lost to more than half a billion.” government department.
to understand the target organization’s As per Kaspersky’s 2015 report D. Setup security and privacy
relationships, activities, interests and (press.kaspersky.com/files/2015/08/ policies properly for each
travel and/or purchasing plans. Kaspersky_Lab_Consumer_Security_ and every social networking
They do this by scraping employee Risks_Survey_2015_ENG.pdf ) . website, you use.
email addresses and other information In 2015, every fourth internet user E. Never forget to use encryption
from the target’s Web site to help make had at least one of their online accounts while sending the emails or
the missives more convincing. In the hacked. This led to unauthorized data across the network.
case where executives or employees messages being sent out in the user’s Conclusion-
have their inboxes compromised by name, and the loss or theft of personal Cybercrime has emerged as a
the thieves, the crooks will scour the data. The study was done on working serious threat, therefore we need to
victim’s email correspondence for employees. Therefore, every single take proper initiatives to curb cross
certain words that might reveal whether personal account hack has the potential border cyber war threats which are
the company routinely deals with wire to grow into a full-scale cyber-attack taking shape. Making people aware
transfers — searching for messages against a business. by imparting training to them or
with key words like “invoice,” “deposit” 2016 was popular for ransom ware conducting workshops can be a very
and “president.” attacks, after cyber crime, hackers ask fruitful technique to curb the cyber
Findings- to pay the ransom in form of bitcoins. crime and provide the social security
After stealing personal information, Safe Guards- to citizens. Government mandate can
cyber criminals gather target’s debit 1. Be careful while sharing personal play a vital role in spreading awareness
card, credit card details, date of birth, information on social networking about cyber crime and preventive
mobile no, PAN number, Voter Id etc. websites. Never disclose Birth measures.
Cyber crime happens due to lack of place, Date of birth or selfies if you References-
cyber-crime awareness. Companies do are on vacation. [1] h ttp : // www.h i n d u sta n t ime s. co m/
not fully realise that protecting assets 2. Install licensed software delhi/man-fined-for-posting-
in the virtual world is more complex 3. Use two way authentication for ex-girlfriend-s-photo/story-
than protecting assets in the physical your email accounts or social I9KGe1mHV6wzr3sAJdk8BK.html
world. networking websites. [2] http://timesofindia.indiatimes.com/
Over Tor network cyber criminals 4. Use OTP (one time password) for city/hyderabad/Cyber-crime-cases-
shoot-up-post-demonetisation/
sell the personal information of people online shopping, money transfer.
articleshow/56129277.cms
after theft. They sell bank’s data, 5. Keep minimal amount in the bank [3] http://economictimes.indiatimes.
corporate data, corporate secrets, account which is linked with online com/news/economy/policy/india-
government secrets and payment is applications or shopping portals. needs-forensic-it-audits-for-
received in bit coins. Moreover they offer 6. Look for “https” and crosscheck banks-to-prevent-card-fraud/
criminal services on hire and rent basis the ownership of the website by articleshow/55002400.cms
too. One can buy drugs, arms using Tor looking at the right top corner of [4] www.wikipedia.com
network. “URL” text box. [5] http://www.verizonenterprise.com/
verizon-insights-lab/dbir/
As per Symantec, spearphishing, 7. Identify Phishing email- As per
which targets humans, was on the Verizon, phishing is the second n

About the Author

Priyanka Tomar is Founder and Director New Era System Pvt Ltd., Certified Ethical Hacker and Cyber Security Professional having 16 years of
experience into Information and Technology field including Software Development, System Analysis, Cyber Security & Forensics. I have written
for some international cyber security magazines and did research work on firewalls and cyber security products for US based company. I am
associated with CBI Academy as visiting faculty, and conducted workshops and seminars on Cyber Security and Cyber Forensics in various
corporate houses, universities like Punjabi University Patiala, Jaipuria Management Institute, JIMS, BSA, Bharti Vidyapeeth College of Engineering
and other educational institutes.
Academics - MCA, PG Diploma in Cyber Security, C|EH (Certified Ethical Hacker), BSc (PCM)

www.csi-india.org
40
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
A R eport

National Conference on
Future Innovations in
Computing Technologies and
Machine Learning (FICTML-17)

National Conference on Future Innovations in Computing University, Rohtak presided. Prof. M. N. Hoda, Director,
Technologies and Machine Learning (FICTML-17) was BVICAM, Bharti Vidyapeeth University, N. Delhi presided over
organised by the Department of Computer Science & the valedictory session.
Applications, M. D. University, Rohtak (Haryana) on 21st
November 2017. The event was indeed a grand success and was attended by
more than 250 delegates. There were 8 Technical Sessions (4
The Chief-Guest was Prof. Yogesh Singh, Vice-Chancellor,
Delhi Technological University, Delhi. Keynote Speaker was Tracks in parallel) chaired by Experts from other universities.
Sh. Vivek Varshney, Managing Director & CEO, VTS Global, There were 111 full-papers and 08 abstracts published in the
Gurgaon. Prof. Bijender K. Punia, Vice-Chancellor, M. D. Conference Proceedings (ISBN 978-93-80544-31-1).

41
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
A R eport

CSI Mumbai Chapter

CSI-IT 2020
Annual Technology Conference and
IT Innovation and Excellence Awards 2017
6-7 October 2017
Theme : Transforming Business with Cognitive Technologies
Director Cognitive Computing & Analytics Practice from IBM
on Day 2.
The general technology sessions were on topics
Cognitive Technologies. The invited talks were on :- 1) Panel
Discussion on : AI and Biometric Technologies by Ms. Gopali
Contractor of Accenture, Mr. Anirudh Kala from Celebal,
Hrishikesh Sherlekar from TCS and Mr. Suresh Shan from
Mahindra Finance 2) Stream Analytics and IoT by Ms. Bharati
Lele, Business Head of Analytics from L & T Infotech 3)
Micro Services Architecture by Mr. Arsh Karaf, CTO from Meta
CSI Mumbai Chapter organised its Annual Technology Magic Global Inc. 4) Robotic Process Automation by Mr. Uday
Conference and IT Innovation and Excellence Awards 2017 in Sanghani, Director, Automation Central Operations Leader
association with (Spoken Tutorials and FOSSEE) IIT Bombay,
title “Transforming Business with Cognitive Technologies” from EY 5) EOS Blockchain Introduction by Mr. Sanjay Mehta,
at Victor Menezes Convention Centre, IIT Bombay, Powai on Director of CORE Media 6) Welcoming the Cognitive Future by
October 6-7 2017 Mr. Devesh Rajadhyax, Founder & CEO of Cerelabs 7) Panel
The key highlight of the conference was IT Innovation and Discussion on : Decision Making Applications in BFSI sector (e.g.
Excellence Awards 2017. 46 winners were awarded in different Risk assessment in Insurance, Loan Approval etc) by Mr. Babu
verticals and industry segments in following categories:- Nair, MD, Publisher & Research Head of Banking Frontiers,
1) Best CIO of the year Dr. Pradeep Pendse, Dean IT of Welingkar, Mr. Ashwin
2) Best CISO of the year Chaudhary from E Com Infotech Ltd. and Mr. Chandrashekhar
3) Best CRO of the year Bhide from Praxis Global Alliance 8) How Computer Vision and
4) Best Start-up Company of the year AI transforming Marketing by Mr. Manoj Shinde, Chief Do’er of
5) Best Cognitive Technology Provider Orbo.ai 9) Industry 4.0 platforms and applications by Mr. Anupam
6) Best Company in terms of Cognitive Technologies Tiwari, Lead – IDC PLM of Accenture 10) AI – Rich Chat Bots by
Implementation Mr. Aakrit Vaish, Co-Founder & CEO of Haptik 11) Hardware
7) Best Company in terms of Information Security loop simulation for multiple vehicles by Prof. Hemendra Arya
Implementation from IIT Bombay 12) AI in Regulatory Compliance by Mr. C P
8) Best Bank in terms of Implementations of Cognitive Srivatsan from Accenture.
Technologies Overall we saw around 300 delegates including award
9) Best Bank in terms of Information Security nominees attending the session and general feedback was
Implementation largerly positive.
10) Best Government Organisation w.r.t. implementation of
Cognitive Technologies Members from the various CSI student branches worked
11) Best Government Organisation implementing Information actively while during the conference.
Security
12) Best Company / Individual for creating Cyber Security
awareness
13) Significant Contributor
The conference inauguration started on Day 1 with
Welcome Address by Mr. Abraham Koshy, Chairman, CSI
Mumbai Chapter and Opening Remark by Prof. Kannan
Moudgalya, Professor, IIT Bombay.
The inaugural address was addressed by Chief Guest Mr.
Nitin Sawant MD, Accenture. The Keynote address on “The
New Cognitive Wave - Driving Business Growth” was given
by Mr. Ashok Pai, VP & Global Head - Cognitive Business
Operations from TCS on Day 1 and by Mr. Kingshuk Banerjee,

www.csi-india.org
42
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 A R eport

Team CSI 2017 in the ExecCom and Review Meeting

held at Kolkata on 9 December 2017
Devaprasanna Sinha
Fellow, Computer Society of India and RVP-II, CSI. E-mail : [email protected]

CSI Communications, the oldest publication of its kind and Treasurer respectively for the current year.
circulated amongst the members of Computer Society of
You will be glad to hear that the Society has been registered
India (CSI) with its name, was published as Volume 1, No.
under the Public Societies Registration Act under certificate
1 in September, 1978. The earlier issues were released as
of registration No. 41 of 1965 with the Registrar of Societies,
Newsletters. The first Newsletter, NEWSLETTER ‘0’ was
Hyderabad.
published with the date 7 February, 1985. It was in the form
of an announcement/letter in five paragraphs from Major The next summer meeting of the meeting will be held at
A Balasubramanian, Secretary. The President was Prof. R Tata Institute of Fundamental Research, Bombay on 4th and
Narasimhan. The address for correspondence was given as c/o 5th June, 1965. This may be taken as notice of meeting. The
Defence Research and Development Laboratory, Hyderabad-5. meeting will devote 2 half days for presentation of papers, one
The contents of this Newsletter were published in earlier half day on invited lectures, and one half day to a business
issues. I may mention only two paragraphs here. meeting.
1. As you may be aware, a General body meeting of the ALL There has been a proposal for staring regional centres of the
INDIA COMPUTER USERS GROUP (AICUG) was held at society at cities that have more than two member institutions.
the IIT, Kanpur on 18 Dec. 1964. It was decided at the We could discuss this at Bombay and perhaps make a start at
meeting to alter the name of the group to the ‘Computer Delhi, Calcutta, Bombay as a first step.
Society Of India’. A Constitution was proposed and duly I also find in this Newsletter, inter alia, the mention about
adopted. The meeting also elected the office bearers Membership with the names of 10 institutions and 2 individuals,
(the President and the Secretary) for the current year. reference to IFIP meeting, details of Summer Meting, proposal
Prof. R. Narasimhan has very kindly accepted to serve as of winter meeting.
President of the Society for the current year.
In Editorial appeared in CSI Newsletter, No. 34, June 1973, the
2. 
Presently the question of Registration of the Society Editorial started with this:
is being progressed. As a prelude to this it seems
necessary that a more formal version of the rules and Communicate or.....
regulations have to be made. A draft is being prepared Living is largely a matter of Communicating. The husband
and would be ready for discussion and approval of the kisses his wife; the customer looks at the price tags; the
next summer meeting being planned for early June. The students raises his hand; the little kid smiles. They all are
venue and further details will be contained in forthcoming communicating. People communicate from morning to night
newsletter. The meeting is being planned for two days out in a modern world. For professional bodies sit is imperative
of which half a day will be devoted to a business session. that its members must communicate with each other.
It may not be out of place to reproduce some of the contents of Conventions and seminars are opportunities for face-to-
the Newsletter No. ‘1’. face communications of professional ideas. The amount of
written and published communication is a yard-stick of the
The society elected Prof. R Narasimhan of the Tata eminence of a scientist these are days. These are research
Institute of Fundamental Research, Bombay and Major A communications.
Balasubramanian of the Defence Research & Development
Laboratory, Hyderabad as the President and Secretary-cum- A lone computer-scientist working in an isolated computer

43
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 A R eport

centre feels that pinch of communication. He has problems over the last few decades in the historical perspective. These
pertaining to work, profession, training or new ideas may trigger interests not only to browse over the textual stuff
to communicate. The CSI journals is one channel of for having nostalgic experiences, but also to look for new
communicating where in research ideas, etc. could be research avenues taking into account the contemporaneity
published. and relevance of those aspects in todays’ world. I know, a few
The Newsletter is a more frequently appearing channel. CSI veterans are writing blogs at regular intervals, putting
There are many columns representing different sectors of oral discussions and histories – structured, semi-structured
professional activity. Members must use this channel more and even unstructured manner in many fora available these
often for communication. No idea is too elementary for this days. Attempts have been made in some/Special issues of CSI
channel. We cannot emphasise the need for communication Communications and also in some Souvenirs published on the
within and outside the Computer Society of India than by occasion of Annual Conventions of CSI and other publications.
repeating cliché ‘communicate or perish’. Special Groups and/or Taskforce have also been created from
CSI Communications/CSI Newsletters with many Newsletters time to time, but one should say that has not yielded the
published from the Chapters from time to time, Proceedings, desired result.
Souvenirs and other publications present a huge repository of The various happenings and attempts, here and there,
articles, papers, photographs, news etc. In other words, along documented or undocumented, have formed the treasure of
with the history, gradual developments of almost all aspects the a source of repository. CSI Warehousing is there – to be
of computer science and engineering, and information augmented on a perennial basis. As is natural, there is another
technology could be traced and generated in diverse ways of scope for exercises of relevant CSI data mining, from time to
hitherto unexplored areas of information technology. In the time. It should bring to the fore many facts and decisions that
above lines, I have reproduced excerpts from only a few of might unfold the new areas, as told earlier.
these. I could produce several other excerpts only from our
earlier CSI Communications, besides other publications, but Lastly I must mention that in the last Execom held at Kolkata, I
due to paucity of space, I am refrained from doing this. But, heard from Mr Sanjay Mohapatra, President, Computer Society
in no uncertain terms, I believe there are ardent readers of India that he could manage to get all the old copies of CSI
for this in this country and also across the world. The time Communications from Mr S Parthasarathy, Fellow and very
has come than ever before to tell others, particularly the senior member of CSI. Under his leadership, CSI has taken a
netxgens, to even document those - the gradual development decision to digitize the all these pages for dissemination of the
of computers science and application, including personalia, same to our members.

Team CSI 2017 in the ExecCom and Review Meeting held at Kolkata on 9 December 2017

www.csi-india.org
44
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
From chapters  &  d i v i s i ons

AMRAVATI CHAPTER

The Chapter organized a Talk on “Cyber Security-Protecting

the Human Point” on 11th December 2017. Mr. Ajay Dubey-
CSI Amravati Chapter, has organized Faculty Development Sr. National Manager – Partner and Alliances (SARRC
Program on Android Application Development in association region) of Force Point made the presentation; it was a well
with Department of Computer Science & Engineering, P. R. attended event with about 65 persons participating; it was a
Pote College of Engineering & Management, Amravati from lively and interactive presentation. Mr. Dubey mentioned that
21st to 25th Nov 2017. The Inaugural ceremony of this Faculty every year the IT industry spends more than the previous
Development Program was held on 21st Nov 2017 which was year on Cyber security but suffers more attacks than the
chaired by Dr. Mrs S. D. Wakde, Principal and inaugurated previous year;while hardware and software securities are
by Dr. G. R. Bamnote, Chairman, CSI Amravati Chapter in important, it is essential that we need to rethink on Cyber
presence of Mr. Amit Dengre, Asst. Commissioner, Amravati Security to make the security effective; in this context, he
Municipal Corporation. Prof. Mohd. Zuhair, Vice-Principal, said that it is necessary for organisations to address the
PRPCEM and the All the HODs of respective departments protection of the most important asset in the Organisation
were also present for this ceremony. The convener of the – which is also the most vulnerable – the Human Point.
Program, Prof. Vijay B. Gadicha welcomed the guests The talk was sponsored by Raksha Technologies Pvt.
and all the participants and gave the brief overview of the Limited-Chennai. IEEE Computer Society, Madras Chapter,
program. Dr. G. R. Bamnote expressed his view about the Hindustan Chamber of Commerce and IEEE Technology and
importance of ongoing technologies such as Android, Cloud Engineering Management Society, Madras also associated
Computing, Big Data. He congratulated department of CSE themselves with CSI-Chennai chapter and members from
for organizing this faculty development program and also these organisations also attended the presentation.
appealed everybody present to join CSI as life member. Mr.
Amit Dengre presented the need of Android Technology in
the Smart city project by Gov. of India. Dr. S. D. Wakde given
warm wishes for this program and quoted the importance
of such activity for the development of faculties. Prof.
Mohd. Zuhair expressed his thoughts regarding need of
crosscutting technologies and how they will be beneficial
in near future and also congratulated CSE department and
CSI for organizing such a beneficial workshop for faculties.
For the inaugural program HODs of respective department,
all the faculties from CSE as well as other departments
were present. The program was coordinated by Prof. F
M Shelke, conducted by Prof. P P Pawade and Prof. K B
Bijwe expressed the vote of thanks. Forty five participants DURGAPUR CHAPTER
participated in the faculty development program.
A lecture meeting was held on 9th October 2017 and attended
by Past and present members of the Computer Society of
India, Durgapur Chapter. The topic of the lecture was “Era of
CHENNAI chapter the Intelligent Enterprise”. The lecture was delivered by Dr.
Chennai Chapter conducted the Guest Lecture on Changing Subrata Dasgupta, Past Chairman and one of the founding
IT Scenarios - Changing Service Options on 18th November members of the Computer Society of India Durgapur
2018 by Mr. Ashok, Founder & CEO of Futurenet Technologies Chapter. Dr. Dasgupta touched upon the role played by
(India) Pvt. Ltd.) the Artificial Intelligence, particularly by the e-commerce

45
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
From chapters  &  d i v i s i ons

companies. He explained how the online consumer buying Ghaziabad Chapter. Keynote speaker Prof. O. P. Verma
behaviour is utilized by various companies through AI delivered a technical talk on Applications of Computational
related software. Earlier, Mr. Gurprit Singh, Past Secretary Intelligence. Dr. Banerjee also lightened our participants
and Treasurer, and DGM I/c (C&IT), DSP welcomed all by sharing his research on Neural Network. E-proceedings
present and Past members and expressed his keen desire to were released by the dignitaries on the occasion. There were
revive the Durgapur Chapter which has not been functioning 53 presentations in four tracks. Prof. Sraban Mukherjee
since 2010 due to some compelling reasons. It may be (Director, IMSEC) presented certificates to the participants.
mentioned that CSI, Durgapur Chapter started functioning Prof Pankaj Agarwal, Program Chair and Dr. Upasana
in 1984. He emphasized the need for a society for the Pandey, Convener of the event welcomed the guests and
computer professionals in the area and promised to make proposed vote of thanks respectively.
the chapter vibrant once again. It was decided that an Ad-
hoc committee would be formed on that day. Dr. Dasgupta
agreed to be the president for the meeting. Two other senior
members and ex-employees of DSP Mr. B Mazumdar, and
Shri Salil Kumar Dutta agreed to approved the committee
formation. Membership forms were distributed in hard copy
to all the potential members and the drive continues after
the meeting too. Mr. Rakesh Ojha, proposed the vote of
thanks. The total number of participants was 35. The Ad-hoc
Committee needs to be formalized by the Computer Society
of India, Mumbai Headquarters. About 15 executives of C&IT
are renewing their membership. All help has been promised VELLORE CHAPTER
by Mr. D Sinha, Regional Vice President (Region II) and Mr.
Subimal Kundu, Nomination Committee member. All the Vellore Chapter organized a two day workshop on “R -
pre-requisites for the start of a chapter were fulfilled. Programming for Engineers” on 9th and 10th Dec’2017
at VIT University. Mr M. Pachayappan, Data Analyst, XiM
Entrepreneurship and Ltd, Chennai, explained the basics
features of R tool, and data mining techniques like linear
regression, classification and clustering. His talk mainly
focused on Web Analytics. Around 65 members participated
in workshop, organized by Dr R Rajkumar, Prof. K Govinda,
Past RVP VII.

GHAZIABAD CHAPTER
The Chapter & IMS Engineering College organized a
National Conference on “Recent Trends in Computational
Intelligence (RTCI-17)” on 11th November 2017. Conference
was inaugurated by Mr. Anil Ji Garg, Chairman, Ghaziabad
Chapter and Mr. Saurabh Agrawal, Patron & Past Chairman

www.csi-india.org
46
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 f rom student branches

Region-I
Sunder Deep Group of Institutions, Ghaziabad DIT University, Dehradun

25-10-2017 – Mr Saurabh Agrawal, Patron & Past Chairman 9-8-2017 - Life Reforming event on WHAT NEXT?
& Mr Anil Ji Garg, Chairman Inaugurated Student Branch
Region-Iii Region-v
The LNM Institute of Information Technology, Jaipur KKR & KSR Institute of Technology & Sciences, Guntur

11-11-2017 & 12-11-2017 - Event on LNMHacks 2.0 29-11-2017 to 30-11-2017 – Event on Internet of
Things Start-up boot camp
Region-v
Prasad V. Potluri Siddhartha Inst. of Tech., Vijayawada Aditya Engineering College, Surampalem

20-11-2017 to 25-11-2017 - STTP on Intelligent Computing & Deep Learning 24-11-2017 - Guest Lecture on Essentials of Network Security by Dr Jeyanthi
G. Narayanamma Institute of Technology and Science, Hyderabad

4-10-2017 - Workshop on Android App Development 24-10-2017 & 25-10-2017 - Workshop on Developing Smart
Applications using IoT

47
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
f rom student branches

Region-v
Lakireddy Bali Reddy College of Engineering, Mylavaram NBKR Institute of Science and Technology, Nellore

13-12-2017 - Workshop on WEB TRENDS by Mr. Sai Satish 22-11-2017 - Training program on Artificial Intelligence Revolution
Anurag Group of Institutions, Hyderabad

17-8-2017 - Seminar on Big Data by Dr. Salman Abdul Moiz 4-12-2017 to 9-12-2017 - Short Term Training program on
HTML 5.0 by Mr. Mruthyunjaya Mendu
Vasavi College of Engineering, Hyderabad

21-10-2017 - Guest Lecture on Computational Geometry by 26-10-2017 - The winners receiving certificate and cash prize
Dr Kishore Kothaplli during Technical Quiz
Vasireddy Venkatadri Institute of Technology, Guntur St Joseph Engineering College, Vamanjoor

21-12-2017 – Event on CodeJam, Potential Professor 7-12-2017 - Industrial visit to Infosys

www.csi-india.org
48
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
f rom student branches

Region-V
B.N.M. Institute of Technology, Bangalore

4-11-2017 & 5-11-2017 – Workshop on Webservices by 4-11-2017 to 6-11-2017 – Workshop on Android Application
Mr. Albin Xavier Development by Mr. Vishwa Kiran
REVA University, Bangalore R. V. College of Engineering, Bangalore

11-12-2017 & 12-12-2017 - Workshop on R-Programming 21-12-2017 to 23-12-2017 - International Conference

CSITSS-2017
Region-V Region-Vi
R.V. College of Engineering, Bangalore Late G N Sapkal College of Engineering, Nashik

21-11-2017 & 28-11-2017 - Workshop on How to Write a 4-11-2017 - Seminar on Intellectual Property Right
Technical Paper by Dr. Shantharam Nayak
Region-Vi
Guru Gobind Singh Polytechnic, Nashik Late G N Sapkal College of Engineering, Nashik

22-12-2017 - Workshop on Personality Development, 4-11-2017 - Seminar on Intellectual Property Right

Etiquettes, Effective Communication by Mr.Vijay Acharya

49
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
f rom student branches

Region-VII
KCG College of Technology, Chennai Kongu Engineering College, Perundurai

24-11-2017 – Workshop on Advanced Photoshop and 16-12-2017 - Interdepartmental Technical Symposium

Dreamweaver Naevus-17

VIT University, Vellore Panimalar Institute of Technology, Chennai

13-12-2017 - Workshop on Python Programming 15-11-2017 - Workshop on Embedded Systems with IoT using
Raspberry Pi

Panimalar Institute of Technology, Chennai Rajalakshmi Engineering College (Autonomous), Chennai

20-11-2017 to 26-11-2017 - Internship Program on Java 15-12-2017 - Workshop on Python

Rajalakshmi Engineering College (Autonomous), Chennai Valliammai Engineering College, Kattankulathur

16-12-2017 - Workshop on IoT and its Applications 23-11-2017 & 24-11-2017 - FDP on EC6504 -
Microprocessor and Microcontroller

www.csi-india.org
50
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 f rom student branches

Region-Vii
Jeppiaar Institute of Technology, Sriperumpudur Valliammai Engineering College, Kattankulathur

18-12-2017 - Training on Phython Programming 18-12-2017 - Training on Phython Programming

Student branches are requested to send their report to

[email protected]
with a copy to
[email protected]

Chapters are requested to send their activity report to

[email protected]
with a copy to
[email protected]

Kindly send High Resolution Photograph with the report.

C S I  C A L E N D A R   2 0 1 7 - 1 8

Gautam Mahapatra, Vice President, CSI, Email: [email protected]

Date Event Details & Contact Information

January CSI Cyber FIF-C(2018) Organised by Mumbai chapter (www.csimumbai.org/cyberfifc/

11-12, 2018 Venue : Victor Menezes Convention Centre, IIT Bombay Contact : [email protected]
Phone No. 022 28235476 / 28235548

January 2nd International Conference on Data Management, Analytics and Innovation (ICDMAI 2018), jointly organized by CSI
19-21, 2018 Division II; CSI Pune Chapter and Audyogik Tantra Shikshan Sanstha’s, Institute of Industrial and Computer Management
and Research, Pune, (IICMR).  http://www.icdmai.org
Contact : Dr. Neha Sharma ([email protected], +91-9923602490)
Dr. Deepali Sawai - [email protected], +91-9921000870

52nd Annual Convention - Organized by CSI Kolkata Chapter - Theme: Social Transformation - Digital Way,
Venue: Science City E-mail: [email protected], [email protected], [email protected]
Website: www.csi2017.in

www.csi-india.org
51
C S I C o m m u n ic a t i o n s | J ANUA R Y 2 0 1 8
 Registered with Registrar of News Papers for India - RNI 31668/1978 If undelivered return to :
Regd. No. MCN/222/2018-2020 Samruddhi Venture Park, Unit No.3,
Posting Date: 10 & 11 every month. Posted at Patrika Channel Mumbai-I 4th floor, MIDC, Andheri (E). Mumbai-400 093
Date of Publication: 10th of every month

CSI Invites Project Proposals from Faculty Members and Students

under the Scheme of R&D Funding for the year 2017-2018
As India’s largest and one of the world’s earliest IT teaching-learning process and evolving a quality control
professional organizations, the Computer Society of India mechanism for academic programmes and curricula.
has always aimed at promoting education and research
ƒƒ Strengthening of industry-institutes interaction through
activities, especially in the advanced technological domains
commercialization of technologies and products
and emerging research fields. It is also committed to take
developed by students and faculty.
the benefits of technological progress to the masses across
India in particular to unrepresented territories. In order to ƒƒ Publication of research studies (ICT penetration,
promote research and innovation meeting the grass-root level technological innovation, diffusion & adaptation), state-
ICT needs and emphasize the importance of joint research by of-the-art reports and case studies of education/ research
faculty-students, the CSI has been providing R&D funding for initiatives.
last several years. ƒƒ Identification of potential new and innovative projects
The CSI Student Branches and member institutions of young faculty, researchers and students for possible
are requested to motivate the young faculty members and business incubation
students (including undergraduate and postgraduate) to
benefit from this scheme. The proposals should have the Indicative Thrust Areas for Research funding
following: aim/objectives, expected outcome, indicative thrust The financial assistance up to ` 10,000/- for hardware/
areas for research funding may be submitted to: Computer software projects would be provided to cover items like
Society of India, Education Directorate, CIT Campus, IV Cross equipment, books/journals, field work, questionnaire,
Road, Taramani, Chennai 600113. Last date for Receipt of computation work and report writing. The indicative thrust
Proposals: 31st January 2018. areas for funding include (but not limited): Technology-
Aims and Objectives OS, Programming Languages, DBMS, Computer &
ƒƒ To provide financial support for research by faculty Communication Networks,Software Engineering, Multimedia
members, especially for developing innovative techniques & Internet Technologies, Hardware & Embedded Systems
and systems to improve teaching-learning and Process & Tools- Requirements Engineering, Estimation
institutional management processes. & Project Planning, Prototyping, Architecture & Design,
Development, Testing & Debugging, Verification & Validation,
ƒƒ To provide financial support to students for developing Maintenance & Enhancement, Change Management,
new systems catering to the needs of socially relevant Configuration Management, Project Management, Software
sectors and/or involving proof of concepts related to Quality Assurance & Process Improvement, Vertical
emerging technologies. Applications- Scientific Applications, Enterprise Systems,
ƒƒ To facilitate interaction / collaboration among Governance, Judiciary & Law Enforcement, Manufacturing,
academicians, practitioners and students. Healthcare, Education, Infrastructure, Transport, Energy,
Defence, Aerospace, Automotive, Telecom, Agriculture &
ƒƒ To develop confidence and core competence among
Forest Management, Inter-disciplinary Applications- CAD/
faculty/students’ through research projects.
CAM/CAE, ERP/SCM, EDA, Geo-informatics, Bioinformatics,
ƒƒ To foster an ambience of ‘Learning by Doing’ and explore Industrial Automation, CTI and Convergence.
opportunities of industry funding and mentoring for Only one proposal is accepted from the Institution.
inculcating professionalism and best practices among
students and faculty. Last date for Receipt of Proposals: 31st January 2018
For further details and application form can be
ƒƒ To recognize innovation and present excellence awards downloaded from CSI portal www.csi-india.org under scrolling
for path-breaking projects through CSI YITP awards and news items as “CSI R&D Projects Application”
industry associations, Govt. Agencies and professional
societies.
Expected Outcome Computer Society of India
ƒƒ Identification of thrust areas, capability assessment, gap Education Directorate,
analysis, recommendations and future education and CIT Campus, IV Cross Road
research directions.
Taramani, Chennai-600113
ƒƒ Integration of research methodologies into the university
Email: [email protected]