Training-Free 3D Editing via Latent Anchor-Aligned Flows
AnchorFlow is a training-free framework for 3D shape editing. It aligns source and target flow trajectories with latent anchors to achieve semantic-consistent, identity-preserving, mask-free 3D editing. (Click the teaser below to see AnchorFlow in action 👇)
This project webpage is built upon the TRELLIS project page. We thank the TRELLIS authors for open-sourcing their webpage design and assets.
If you find our work useful, please consider citing:
@article{zhou2025anchorflow, title={AnchorFlow: Training-Free 3D Editing via Latent Anchor-Aligned Flows}, author={Zhou, Zhenglin and Ma, Fan and Gui, Chengzhuo and Xia, Xiaobo and Fan, Hehe and Yang, Yi and Chua, Tat-Seng}, journal={arXiv preprint arXiv:2511.22357}, year={2025}, }