KinePose: A temporally optimized inverse kinematics technique for 6DOF human pose estimation with biomechanical constraints

• Published in: arXiv.org
• Main Authors: Gildea, Kevin, Mercadal-Baudart, Clara, Blythman, Richard, Smolic, Aljosa, Simms, Ciaran
• Format: Paper Journal Article
• Language: English
• Published: Ithaca Cornell University Library, arXiv.org 06.09.2022
• Subjects: Accuracy; Algorithms; Ambiguity; Biomechanics; Computer Science - Computer Vision and Pattern Recognition; Computer vision; Constraints; Errors; Inverse kinematics; Joints (anatomy); Kinematics; Knee; Machine learning; Mathematics - Optimization and Control; Optimization; Optimization techniques; Pose estimation; Quadratic programming; Solution space; Three dimensional models; Three dimensional motion
• ISSN: 2331-8422
• DOI: 10.48550/arxiv.2207.12841

Computer vision/deep learning-based 3D human pose estimation methods aim to localize human joints from images and videos. Pose representation is normally limited to 3D joint positional/translational degrees of freedom (3DOFs), however, a further three rotational DOFs (6DOFs) are required for many po...