Portfolio

SignLangNET

SignLangNET is a project aimed at interpreting sign language gestures using deep learning techniques. It utilizes the MediaPipe library for hand and body pose estimation and employs Long Short-Term Memory (LSTM) networks for sequence modeling, with a 96.4% gesture translation accuracy.

ClusterCraft

An intuitive Flask web app employing a custom k-Means algorithm for image clustering, deployed on a flask app. The project attracted 3 contributors during Hacktoberfest event, fostering open-source engagement and skill enhancement for new contributors.

Multi-Physics Simulation Engine

A physics simulation engine written completely in python, to simulate fluids, molecular dynamics, gravitation, and other physical phenomena.