1. Which of the following is the main purpose of evaluating a model?

a) To increase the size of the dataset

b) To check how well the model performs
c) To add more features to the model
d) To visualize the dataset
Answer: b) To check how well the model performs
2. In AI, what does ‘accuracy’ of a model represent?
a) Number of wrong predictions
b) Percentage of correct predictions
c) Time taken to train the model
d) The total number of data points
Answer: b) Percentage of correct predictions
3. What is a confusion matrix used for?
a) To visualize the dataset
b) To summarize the performance of a classification model
c) To clean the data
d) To build the model
Answer: b) To summarize the performance of a classification model
4. Which metric is most suitable for imbalanced datasets?
a) Accuracy
b) Precision and Recall
c) Training Time
d) Data Distribution
Answer: b) Precision and Recall
5. What is the formula for Accuracy?
a) (TP + TN) / (TP + TN + FP + FN)
b) TP / (TP + FP)
c) TP / (TP + FN)
d) FP / (TP + FN)
Answer: a) (TP + TN) / (TP + TN + FP + FN)

6. If a spam detection model marks 80 out of 100 spam emails correctly but marks
10 genuine emails as spam, what is its precision?
a) 80 / (80 + 10) = 0.89
b) 80 / 100 = 0.8
c) 10 / (80 + 10) = 0.11
d) 90 / 100 = 0.9
Answer: a) 0.89
7. A model gives high accuracy but poor performance on new data. What problem
does this indicate?
a) Underfitting
b) Overfitting
c) Bias
d) High variance
Answer: b) Overfitting
8. If a model never predicts a positive class, which evaluation metric will be most
misleading?
a) Precision
b) Recall
c) Accuracy
 9.
d) F1-Score
Answer: c) Accuracy
10. In a hospital model that predicts whether a patient has a disease, which is more
important: Precision or Recall? Why?
a) Precision, because false positives are dangerous
b) Recall, because missing actual patients (false negatives) is dangerous
c) Accuracy, because all predictions must be correct
d) None of the above
Answer: b) Recall
11. What does an F1-score of 1 represent?
a) Perfect balance of precision and recall
b) Low performance model
c) High number of false negatives
d) Only precision is high
Answer: a) Perfect balance of precision and recall

🔹 Critical Thinking / Real-life Based Questions

11. You develop an AI model to detect fraud transactions. The model gives 99%
accuracy but misses most frauds. What does this indicate?
a) Model is perfect
b) Data is balanced
c) Model is biased towards non-fraud class
d) Model is overfitting
Answer: c) Model is biased towards non-fraud class
12. If your model predicts COVID positive for all patients and data is 90% positive,
what will be the accuracy?
a) 0%
b) 10%
c) 90%
d) 100%
Answer: c) 90%
13. Why is cross-validation important during evaluation?
a) It helps test the model on different subsets of data
b) It increases dataset size artificially
c) It removes missing data
d) It reduces the number of features
Answer: a) It helps test the model on different subsets of data
14. Which of the following situations shows underfitting?
a) Model performs poorly on training data and test data
b) Model performs well on training data but poorly on test data
c) Model performs well on both
d) Model memorizes the dataset
Answer: a) Model performs poorly on training data and test data
15. Why should we not rely only on accuracy as a metric?
a) Accuracy is difficult to calculate
b) Accuracy ignores class imbalance and misclassification cost
 c) Accuracy doesn’t depend on predictions
d) Accuracy increases model complexity
Answer: b) Accuracy ignores class imbalance and misclassification cost

You’re very welcome 😊

Here’s a complete and well-balanced set of MCQs for Chapter 5 – Computer Vision (CV)
from Class 10 AI (CBSE – AI 417), aligned with the CBSE curriculum and containing
conceptual, application-based, and HOTS (Higher Order Thinking Skills) questions —
all with answers and brief reasoning.

🤖 Chapter 5 – Computer Vision (CV): MCQ Question

Bank
🔹 Basic Understanding (1 mark each)

1. What is the main goal of Computer Vision?

a) To make computers see and understand images or videos
b) To process text documents
c) To generate audio output
d) To manage databases
Answer: a) To make computers see and understand images or videos

2. Which of the following is not an application of Computer Vision?

a) Face recognition
b) Object detection
c) Speech recognition
d) Image classification
Answer: c) Speech recognition

3. Which step usually comes first in the Computer Vision pipeline?

a) Feature extraction
b) Image acquisition
c) Model evaluation
d) Segmentation
Answer: b) Image acquisition

4. What does an image consist of?

a) Only colors
b) Pixels arranged in rows and columns
c) Audio signals
 d) Video frames
Answer: b) Pixels arranged in rows and columns

5. Which color model is commonly used to represent digital images?

a) RGB
b) CMYK
c) HSV
d) Both a and c
Answer: d) Both a and c

🔹 Application & Understanding

6. In image preprocessing, converting a colored image into black and white is called
—
a) Color enhancement
b) Grayscale conversion
c) Segmentation
d) Thresholding
Answer: b) Grayscale conversion

7. Which of the following libraries is used for CV in Python?

a) NumPy
b) OpenCV
c) Matplotlib
d) Pandas
Answer: b) OpenCV

8. If you want to blur an image to remove noise, which technique will you apply?
a) Edge detection
b) Smoothing filter
c) Thresholding
d) Histogram equalization
Answer: b) Smoothing filter

9. Which function of OpenCV is used to read an image?

a) cv2.show()
b) cv2.imread()
c) cv2.display()
d) cv2.load()
Answer: b) cv2.imread()
 10. In Computer Vision, segmentation helps in —
a) Combining multiple images
b) Dividing an image into meaningful regions
c) Reducing image size
d) Increasing pixel value
Answer: b) Dividing an image into meaningful regions

🔹 HOTS / Critical Thinking

11. A self-driving car detects a pedestrian and stops. Which AI domain makes this
possible?
a) NLP
b) Computer Vision
c) Robotics only
d) Data Science
Answer: b) Computer Vision

12. Why is preprocessing necessary before feeding images to a model?

a) To increase file size
b) To reduce complexity and standardize data
c) To blur the image
d) To add more noise
Answer: b) To reduce complexity and standardize data

13. An AI model detects faces correctly in bright light but fails at night. Which
improvement is needed?
a) Add more training data in varied lighting
b) Decrease image resolution
c) Remove preprocessing
d) Convert to grayscale only
Answer: a) Add more training data in varied lighting

14. If you want to highlight the boundaries of objects, which technique should be
used?
a) Edge detection
b) Histogram equalization
c) Color enhancement
d) Image resizing
Answer: a) Edge detection
 15. Which step ensures that the model focuses only on the region of interest in an
image?
a) Cropping or Segmentation
b) Scaling
c) Grayscale conversion
d) Augmentation
Answer: a) Cropping or Segmentation

🔹 Real-life Scenario / Case based (HOTS)

16. An airport uses cameras to verify passengers’ identities automatically. Which

CV task is this?
a) Image classification
b) Face recognition
c) Object tracking
d) Optical character recognition
Answer: b) Face recognition

17. A traffic monitoring system counts vehicles passing through a toll booth. This is
an example of —
a) Object detection and tracking
b) NLP
c) Sentiment analysis
d) Data visualization
Answer: a) Object detection and tracking

18. A mobile app converts handwritten text to digital text. Which technology is being
used?
a) Speech recognition
b) OCR (Optical Character Recognition)
c) Machine translation
d) Computer graphics
Answer: b) OCR (Optical Character Recognition)

19. In facial-recognition attendance systems, why must lighting and camera angles
be controlled?
a) To reduce bias and misidentification
b) To make the system slower
c) To avoid privacy issues
 d) To decrease accuracy
Answer: a) To reduce bias and misidentification

20. A CV model identifies a banana as a mango due to similar color. What is the
model’s error type?
a) False positive
b) False negative
c) True positive
d) Overfitting
Answer: a) False positive

Would you like me to also prepare a Question Bank PDF for Chapter 5 (Computer Vision)
with:

 💡 MCQs (with answers)

 🧩 Assertion–Reason questions
 📘 Case study-based questions
formatted in CBSE paper style (for classroom/assessment use)?