CVPR2025

FreeCloth: Free-form Generation Enhances Challenging Clothed Human Modeling

Hang Ye, Xiaoxuan Ma, Hai Ci, Wentao Zhu, Yizhou Wang

Abstract

Generation (a) (b) Replication Part Seg. POP FITE Ours Figure 1. (a) An overview of our framework for modeling clothed humans. Based on the specific modeling needs of different regions, we employ a dedicated strategy to handle various clothing areas. Specifically, for loose regions (green) that are less affected by body movements and require more freedom, we propose free-form generation to enhance flexibility. For near-body clothing areas (blue), we apply LBS-based deformation, while unclothed regions (yellow) that do not require deformation can be directly replicated. (b) Visual comparison between prior arts (POP [39], FITE [33]) and our method on challenging clothing. Our method captures more highfidelity details and achieves superior visual quality and realism. Code is available at https://alvinyh.github.io/FreeCloth .