2D Human Pose Estimation
47 papers with code • 5 benchmarks • 17 datasets
What is Human Pose Estimation? Human pose estimation is the process of estimating the configuration of the body (pose) from a single, typically monocular, image. Background. Human pose estimation is one of the key problems in computer vision that has been studied for well over 15 years. The reason for its importance is the abundance of applications that can benefit from such a technology. For example, human pose estimation allows for higher-level reasoning in the context of human-computer interaction and activity recognition; it is also one of the basic building blocks for marker-less motion capture (MoCap) technology. MoCap technology is useful for applications ranging from character animation to clinical analysis of gait pathologies.
We start from a high-resolution subnetwork as the first stage, gradually add high-to-low resolution subnetworks one by one to form more stages, and connect the mutli-resolution subnetworks in parallel.
HigherHRNet even surpasses all top-down methods on CrowdPose test (67. 6% AP), suggesting its robustness in crowded scene.
We demonstrate that our pose-based framework can achieve better accuracy than the state-of-art detection-based approach on the human instance segmentation problem, and can moreover better handle occlusion.
We introduce associative embedding, a novel method for supervising convolutional neural networks for the task of detection and grouping.
We rethink a well-know bottom-up approach for multi-person pose estimation and propose an improved one.