Deep neural network trained to detect eye contact from facial image

Head Pose Estimator on Caffe

Real-time Face & emotion recognition system using a lightweight CNN, with OpenCV webcam inference, GroqCloud for chatbot & FastAPI + frontend web deployment.

A real-time facial analysis platform built with Flask, OpenCV, TensorFlow, PyTorch, and Next.js, featuring live face detection, age & gender estimation, and emotion recognition. Designed for robotics club events, tech fairs, and interactive AI demos, with a futuristic cyberpunk UI powered by Arwes.

OneStopVision is an open-source toolkit offering a comprehensive suite of algorithms for face and body analysis, landmark extraction, and ControlNet integration in Stable Diffusion.

DrowsinessGuard: A real-time drowsiness detection system that enhances driver safety using computer vision and facial landmark analysis. Features precise eye tracking, face direction monitoring, and yawn detection to alert drivers of fatigue, helping prevent accidents caused by drowsy driving.

PROSOPO is a comprehensive facial recognition system designed with a focus on fairness and bias mitigation. It integrates state-of-the-art detection and recognition models with advanced fairness-aware techniques to ensure equitable performance across different species.s.

AI-powered facial engagement analyzer for monitoring attention via webcam using MediaPipe and FastAPI.

facialAnalysisLiteDetection2D: Detector implementation for the analysis of facial features. Using the cvlib API analysis for gender extraction and prediction. Via OpenCV DNN, age prediction. Through tensorflow, emotion detection. Image source reception via YARP, processing and sending of detections via YARP port.

A deep learning-based system for detecting deepfakes in images and videos, with integrated facial emotion analysis. Built using EfficientNetB4, CNN-GRU, OpenCV, and DeepFace to evaluate both authenticity and emotional expression across video frames. Supports detailed evaluation metrics and real-time graph visualization.

AI clinical decision support system for automated facial palsy assessment using computer vision

AI-powered facial skin quality analysis — 7-zone scoring, concern heatmaps, biological age estimation. Upload a selfie, get instant results.

Official codebase for the paper "Synthetic Data for Fairness: Bias Mitigation in Facial Attribute Recognition" (IEEE, 2025). Includes an automated pipeline for generating, labeling, and evaluating synthetic facial datasets under demographic control using SD/GAN + fairness metrics.

Multimodal Parkinson's disease analysis from YouTube videos — facial (hypomimia) and speech (dysarthria) feature extraction with a late-fusion deep learning model.

Real-time facial emotion recognition using OpenCV and deep learning with webcam and IP camera support.