Abstract:In the domain of 3D Human Pose Estimation, which finds widespread daily applications, the requirement for convenient acquisition equipment continues to grow. To satisfy this demand, we set our sights on a short-baseline binocular setting that offers both portability and a geometric measurement property that radically mitigates depth ambiguity. However, as the binocular baseline shortens, two serious challenges emerge: first, the robustness of 3D reconstruction against 2D errors deteriorates; and second, occlusion reoccurs due to the limited visual differences between two views. To address the first challenge, we propose the Stereo Co-Keypoints Estimation module to improve the view consistency of 2D keypoints and enhance the 3D robustness. In this module, the disparity is utilized to represent the correspondence of binocular 2D points and the Stereo Volume Feature is introduced to contain binocular features across different disparities. Through the regression of SVF, two-view 2D keypoints are simultaneously estimated in a collaborative way which restricts their view consistency. Furthermore, to deal with occlusions, a Pre-trained Pose Transformer module is introduced. Through this module, 3D poses are refined by perceiving pose coherence, a representation of joint correlations. This perception is injected by the Pose Transformer network and learned through a pre-training task that recovers iterative masked joints. Comprehensive experiments carried out on H36M and MHAD datasets, complemented by visualizations, validate the effectiveness of our approach in the short-baseline binocular 3D Human Pose Estimation and occlusion handling.

Simple Baselines for Human Pose Estimation and Tracking

X-HRNet: Towards Lightweight Human Pose Estimation with Spatially Unidimensional Self-Attention

Simple and Lightweight Human Pose Estimation

A Simple Baseline for Pose Tracking in Videos of Crowded Scenes

RSB-Pose: Robust Short-Baseline Binocular 3D Human Pose Estimation with Occlusion Handling

FollowMeUp Sports: New Benchmark for 2D Human Keypoint Recognition

A Simple Baseline for Pose Tracking in Videos of Crowed Scenes

SIMPLE: SIngle-network with Mimicking and Point Learning for Bottom-up Human Pose Estimation

Towards Simple and Accurate Human Pose Estimation with Stair Network

Bottom-Up Human Pose Estimation by Ranking Heatmap-Guided Adaptive Keypoint Estimates

A Top-down Approach to Articulated Human Pose Estimation and Tracking

AlphaPose: Whole-Body Regional Multi-Person Pose Estimation and Tracking in Real-Time

ViTPose: Simple Vision Transformer Baselines for Human Pose Estimation

Recent Advances in 3D Human Pose Estimation: From Optimization to Implementation and Beyond

Rethinking on Multi-Stage Networks for Human Pose Estimation

Human Pose Estimation Based on Lightweight Multi-Scale Coordinate Attention

Low-Budget Real-time Pose Estimation Method for Personal Physical Exercise Assistant

RePose: Learning Deep Kinematic Priors for Fast Human Pose Estimation

Research on Human Posture Estimation Algorithm Based on YOLO-Pose

Robust Estimation of 3D Human Poses from a Single Image

Single Person Pose Estimation: A Survey