The main goal of this repository is to share a variety of deep learning projects explained in a friendly manner, taking the approach of create any project from scratch and follow a step-by-step building of all the components of a specific project. I am working primarly in the development of Natural Language Processing and Computer Vision applications, and I consider that this repository is in active development, as I'm going to frequently load new projects. Next, you can find an overview of the projecst that currently compose this repo. Please feel free to use them if you find them valuable for you own developments.

For this type of applications, I'm completely focused on the use of Large Language Models (LLMs). At this moment, you can find a Sentiment Analysis project that involves the use of DistillBERT and the implementation of the strategy of data augmentation in text using the TextAttack. PyTorch is used to conduct the training and evaluation stages, whereas Hugging Face is used to load DistillBERT and its corresponding tokenizer, as well as to load the emotion dataset.

Currently, you can find a Thermal Face Recognition System implemented through Tensorflow/Keras and using the Terravic Facial IR Database. For this project, I performed the training stage using a Convolutional Neural Network and the strategy of fine-tunig.