Projects sharing features with MediaPipePyTorch

The project is an official implement of our ECCV2018 paper "Simple Baselines for Human Pose Estimation and Tracking(https://arxiv.org/abs/1804.06208)"

MMPose is a PyTorch-based pose estimation toolbox and deep learning training pipeline designed for detecting 2D and 3D keypoints on humans, animals, and faces. It serves as a computer vision model zoo and a framework for both 2D pose estimation and 3D pose lifting. The project is distinguished by its modular architecture and extensibility, employing a registry-based system and hierarchical configurations to allow for custom algorithm integration and model pipeline customization. It supports diverse estimation paradigms, including top-down, bottom-up, and two-stage pose lifting workflows. The

This is a multi-person pose estimation framework designed for real-time human keypoint detection. It functions as a bottom-up human pose estimator that identifies skeletal joints across all people in a scene without requiring a separate person detector. The system utilizes a convolutional neural network model to generate heatmaps and vector fields for posture analysis. It specifically implements part affinity fields to encode the location and orientation of limbs, allowing the model to connect individual joints into complete skeletons. The project covers computer vision motion analysis and d