HairDeck Generates Animation-Ready Stylized Hair from One Image
An arXiv paper presents HairDeck, a generative system that turns a single image into stylized hair cards and the spring-bone chains that animate them. The method represents each hair card as one token and uses a set diffusion transformer to denoise all cards jointly. The authors report that, on CHARM's test set, HairDeck lowers Chamfer distance by 44%, reduces card-orientation error by more than half, and runs 33 times faster; they also report better results than CHARM, image-to-3D pipelines, and strand-based methods from single images. These benchmark results are author-reported.
HairDeck provides a single-image workflow for generating animatable stylized-hair assets that can export to downstream formats. The authors report substantial gains over CHARM, image-to-3D pipelines, and strand-based methods, although the results require independent validation.
Computer Science > Computer Vision and Pattern Recognition
arXiv:2610.40599 (cs)
[Submitted on 29 Oct 2026]
Title:HairDeck: Animation-Ready Stylized Hair Cards from a Single Image
Authors:Yuze He, Yixiang Wei, Wang Zhao, Chuhao Chen, Yanning Zhou, Yong-Jin Liu
View a PDF of the paper titled HairDeck: Animation-Ready Stylized Hair Cards from a Single Image, by Yuze He and 5 other authors