Quadruped robot pose estimated accurately from video sequences
Camera-to-Quadruped Robot Pose Estimation from Sequential Images
RoboticsComputer Vision and Pattern Recognition
Summary
Tracking the exact position and joint movement of four-legged robots is hard when seen from normal cameras because their legs can block each other and look similar. To solve this, the authors created QuadPose, a method that estimates the robot’s full body pose from regular video. They also made QuadSyn, the first dataset to test how well such methods work. Their system works quickly and accurately enough to help monitor these robots in dangerous places.
What this means in practice
- •For robotics operators: Monitor and control quadruped robots remotely in hazardous environments using accurate video-based pose estimation.
- •For industrial inspection teams: Improve safety and reliability of facility inspections by tracking robotic assistants’ movements precisely via standard cameras.
Authors
Seobin Hwang, Ji-Seong Beom, Yeong-Jun Cho
Abstract
Quadruped robots are increasingly deployed in hazardous environments such as disaster response and facility inspection, where external-camera monitoring is essential for safe operation. This setting is challenging because quadrupeds are articulated systems with time-varying leg joints, leg symmetry, and frequent self-occlusions, which make keypoint correspondence ambiguous. Moreover, no public benchmark exists, making quantitative comparison difficult. To address these challenges, we propose QuadPose, which estimates the full-body 18-DoF pose (6-DoF base pose + 12-DoF joint angles) from monocular video sequences, and introduce QuadSyn, the first camera-to-quadruped robot pose benchmark. On QuadSyn, QuadPose achieves 92.18% PCK@2.5px, 95.54% PCK-AUC, 85.94% ADD@20cm, 66.98% ADD-AUC, and 3.65 degree mean joint-angle error at approximately 33 FPS.