Khoi Tran Dinh Anh

Khoi Tran Dinh Anh

AI Learning

Face Analysis

Developed an advanced Human Face Analysis system utilizing YOLOv8 for real-time face detection and a customized ResNet model for in-depth facial feature analysis. This project aimed to accurately detect, track, and analyze various facial expressions, emotions, and facial attributes in dynamic environments.

  • Face Detection with YOLOv8: Fine-tuned a YOLOv8 model to detect human faces with high precision and efficiency in real-time, even in challenging conditions such as varying lighting, occlusions, and crowded scenes.

  • Custom ResNet for Feature Analysis: Modified a pre-trained ResNet architecture to analyze specific facial attributes such as age, gender, emotion, and facial landmarks. The model was further optimized to classify and recognize subtle variations in facial expressions.

  • Real-time Facial Expression Tracking: Implemented tracking algorithms to follow facial expressions over time, providing dynamic, continuous analysis of emotional states and behaviors.

  • Applications: The system can be applied in various fields such as security, marketing (customer emotion analysis), human-computer interaction, and medical diagnosis for facial expression recognition.

  • Technologies Used: Python, OpenCV, YOLOv8, PyTorch, ResNet, real-time image processing algorithms.

This project highlights expertise in deep learning, facial analysis, and the application of computer vision technologies for real-time, practical use cases in both academic and industrial settings.