This project presents a real-time human pose detection system using deep learning techniques to identify
body keypoints from images and video streams. The system utilizes YOLOv8 Pose along with computer
vision libraries such as OpenCV to accurately detect and track human body movements in real time. The
implementation is integrated into a Streamlit-based web application, providing an interactive interface
where users can perform pose estimation using a live camera or uploaded images and videos. The system
processes input data efficiently and visualizes keypoints and skeletal structures for better understanding
of human posture. Furthermore, the proposed solution supports posture analysis and can be applied in
domains such as fitness monitoring, rehabilitation, sports analysis, and human activity recognition. The
model ensures high accuracy, low latency, and ease of use, making it suitable for real-world applications
and scalable for future enhancements.
