AI/ML Free (open-source)
TensorFlow Pose Estimation (PoseNet)
Pre-trained deep learning model for real-time human pose estimation from images and videos.
Overview
TensorFlow’s PoseNet model detects 17 key body points (e.g., nose, shoulders, elbows, wrists, hips, knees, ankles) from a single image or video frame. It runs efficiently on GPU and is suitable for batch processing in Colab.