Projects

Visual Semantic Embedding (VSE) models are one of those models which combine the fields of Computer Vision & NLP together. This project takes its core ideas from VSE models but simplifies it. We learn joint embeddings for images as well as text. Check it out for yourself!

Part of my Diploma thesis. This project is on end - to - end classification of Dry Beans. Every step from EDA, Feature Engineering, Selection to Deployment has been done in details with documents for each phase. Check it out!

This is a PyTorch implementation of CheXNet paper which was part of my internship. 

Complete pipeline to train a model for Hindi Character Recognition including both Vyanjans (letters) & Numbers. There is a streamlit as well as a gradio app for you to check it out. The implementation is both in PyTorch-Lightning and from scratch PyTorch.

This contains implementations of the original UNet and Attention UNet from scratch in PyTorch