a GUI application, which uses YOLOs (YOLOv8, YOLO11, YOLOv13) for Object Detection/Tracking, Human Pose Estimation/Tracking from images, videos or camera

YoloX with tracking for a bare Raspberry Pi 4 using ncnn.

NanoDet with tracking for a bare Raspberry Pi 4 using ncnn.

ByteTrack for multi-object tracking with many Detectors

This project provides an intelligent real-time traffic monitoring system that detects, tracks, counts vehicles, and estimates speed using perspective transformation.

Counting Bees

Real-time soccer player tracking & re-identification with YOLOv11, plus a benchmarked comparison of ByteTrack, BoT-SORT, DeepSORT, and Norfair on 720p/4K footage. Python · OpenCV · Ultralytics.

Real-time CCTV threat detection pipeline built with Python, YOLO26n, ByteTrack, MediaPipe Pose, and OpenCV to detect fights, loitering, weapons, and abandoned objects from video streams.

AI-powered retail intelligence system that converts CCTV footage into real-time store analytics using YOLOv8, ByteTrack, FastAPI, and live dashboards.

A self-contained ByteTrack implementation extracted from Ultralytics YOLO. This tracker works with any detector and has no Ultralytics runtime dependency.

An attendance solution that recognizes students from classroom video and automates accurate, contactless attendance tracking.

Real-time Store Intelligence System that transforms CCTV footage into actionable retail insights through computer vision, event processing, and analytics dashboards.