PHAF-Personalized Hand Avatars in a Flash
2026-06-02 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors developed PHAF, a method to create realistic 3D hand models from just two photos of a person's hand. Their technique is much faster than previous methods, producing detailed textures quickly so it can work in real-time on devices like phones or AR glasses. They use smart mesh alignment and texture refinement to make the hand look smooth and accurate from different angles. This system also works well with existing graphics software and captures hand movements realistically.
personalized avatarphoto-realistic renderingmesh alignmenttexture extractionview-based inpaintingparametric hand modelmulti-view renderingreal-time graphicsAR/VR applicationstexture generation
Authors
Meghana Shankar, Akanxit Upadhyay, Anmol Namdev, Green Rosh KS, Pawan Prasad BH
Abstract
We present PHAF-Personalized Hand Avatars in a Flash, a personalized photo-realistic hand avatar which provides high quality multi-view renders from just two images (dorsal and palmar views).Unlike slow optimization-based techniques, PHAF generates fast personalized textures for real-time deployment on edge devices. Our approach combines semantic guided mesh alignment and densified texture extraction to transfer high-frequency details efficiently. A view-based inpainting network refines textures ensuring smooth, continuous appearance. PHAF generalizes to novel viewpoints and leverages a parametric hand model for accurate articulations, making it compatible with standard graphics engines. Experiments show it is comparable to existing methods in visual fidelity while drastically reducing texture generation time by 30 times, enabling practical AR/VR applications.