Projects

This section showcases some of the key projects I have worked on, highlighting my skills in Computer Vision, Deep Learning, and AI solutions.

Anchor Boxes Initialization with KMeans for Faster-RCNN

Improved object detection performance by implementing KMeans clustering for optimal anchor box initialization in Faster-RCNN. This approach reduced convergence time by 50% and increased mAP by 6% on custom datasets.

TensorFlow Computer Vision Object Detection
View on GitHub

Vehicle Detection and Tracking

A complete pipeline for real-time vehicle detection and tracking in video streams using YOLOv5, OpenCV, and PyTorch. Includes multi-object tracking and visualization capabilities.

PyTorch YOLOv5 OpenCV Tracking
View on GitHub

Retail Shelf Empty Space Detection

Computer vision solution for detecting empty spaces in retail store shelves and racks. Helps optimize inventory management and improve restocking efficiency in retail environments.

Computer Vision Python Deep Learning Retail Analytics
View on GitHub