My Projects

Real-time collaborative drawing app with WebSocket-powered chat, built using Next.js, React, and Prisma.

AI system for detecting heart and lung anomalies from auscultation sounds with a clinical interface.

A second-brain-style full stack knowledge management system built with MERN stack + JWT auth.

A deep learning system to identify ukulele chords using CNNs and audio feature extraction with Librosa, deployed via Streamlit.

Multiclass NLP model for classifying sentences in medical research abstracts using Universal Sentence Encoder and TensorFlow.

Medical image segmentation pipeline using nnU-Net, UNETR, MiSCNN, and TotalSegmentator with 3D volume processing.

Hackathon project using satellite image classification to determine cloud coverage levels from satellite images.

Biometric recognition system based on palm prints with full paper replication and dataset preprocessing.

A TensorFlow model to classify Capuchin bird calls using audio signal processing and spectrograms.

Used YOLO and OpenCV to detect and count cars in traffic surveillance footage.

Real-time footfall analysis project that detects and counts people in shopping malls using object detection.

Detection of personal protective equipment like helmets, vests, and masks in video feeds for workplace safety.

Real-time detection of poker cards and hands using OpenCV and object detection.

Regression model to predict bulldozer resale prices using ensemble ML methods.

TensorFlow model that classifies dog breeds from images using transfer learning and fine-tuning.

Cryptocurrency price forecasting model using time series techniques with TensorFlow.

Reproduced Vision Transformer (ViT) as per original paper using PyTorch. Trained on image dataset for classification.

A full-featured todo app with JWT auth, Node.js backend, and CRUD operations.