Towards Geometric-Photometric Joint Alignment for Facial Mesh Registration

arXiv:2403.02629v1 Announce Type: cross
Abstract: This paper presents a Geometric-Photometric Joint Alignment(GPJA) method, for accurately aligning human expressions by combining geometry and photometric information. Common practices for registering human heads typically involve aligning landmarks with facial template meshes using geometry processing approaches, but often overlook photometric consistency. GPJA overcomes this limitation by leveraging differentiable rendering to align vertices with target expressions, achieving joint alignment in geometry and photometric appearances automatically, without the need for semantic annotation or aligned meshes for …

abstract alignment arxiv cs.cv cs.gr geometry human information mesh meshes paper practices processing registration template type