Autonomous Driver AI is a production-grade simulation engine designed to test autonomous vehicle logic in a 2D environment. Built using FastAPI and NumPy, the system leverages a kinematic bicycle model and raycasting sensors to simulate real-world vehicle dynamics and obstacle detection at a consistent 60 FPS.

This project is a real-time gesture recognition system designed to control arcade-style games (like Subway Surfers) using hand movements. Built with OpenCV and MediaPipe, the application tracks the user's index finger via webcam and maps its position to keyboard inputs.

An AI-powered project where a simulated robot uses 270-degree LiDAR and Reinforcement Learning to autonomously navigate mazes, optimizing real-world logistics like hospital delivery and warehouse automation.

SignSpeak AI translates sign language into real-time speech using computer vision. This accessible tool bridges communication gaps, empowering non-verbal individuals to express needs and interact effortlessly in any daily environment.

It is an automated traffic signal control system that uses real-time data to optimize intersection throughput. By integrating YOLOv8 for vehicle detection and a dynamic Decision Engine, the system replaces static timers with "smart" signals that adapt to actual road conditions.

It is a touchless navigation system using Python and Mediapipe. It allows surgeons to control digital X-rays via hand gestures, maintaining a sterile environment by replacing mice with mid-air finger tracking.

An interactive simulator visualizing Pathfinding AI and Procedural Generation. It demonstrates how autonomous agents navigate complex environments using BFS, DFS, and heuristic-based logic to solve randomly generated, real-time mazes.

SentinelAI is a proactive surveillance backend that utilizes computer vision to identify security risks in real-time. Unlike traditional passive monitoring, this system calculates a dynamic Threat Score based on the presence of specific objects, individuals, and facial detection, triggering automated alerts when thresholds are breached.

NeuralFace is a sophisticated real-time biometric monitoring system that uses computer vision to quantify human emotional states and stress levels. By leveraging MediaPipe’s 468-point face mesh, the engine extracts subtle micro-expressions and translates them into actionable psychological metrics.

This Multi-Object Smart Scanner uses Llama 4 Scout and Groq to provide real-time visual checkout. It instantly identifies, counts, and prices multiple items from one photo, eliminating manual barcode scanning.

An interactive digital companion that blends high-fidelity rendering with conversational AI. Engage in real-time dialogue with a lifelike character designed to listen, respond, and animate.