TanGO: Training-Free 3D Editing via Tangent-Space Guidance and Optimization
Computer Vision and Pattern Recognition
Summary
The authors present TanGO, a new method to improve 3D model editing without retraining the model. Unlike previous methods that treat all parts of the model the same and cause weird distortions, TanGO changes each part individually using a math approach based on how the model generates shapes. This selective control reduces errors and makes edits look better. Their experiments show TanGO works better than other current editing methods.
Authors
Siwoo Lim, Sunjae Yoon, Gwanhyeong Koo, Hyeonseo Yun, Chang D. Yoo
Abstract
While recent flow-matching 3D generative models (e.g., VecSet) adopt structured representations, their tokens share global context, causing conventional training-free editing to suffer from semantic artifacts such as collapsed preserved regions or incomplete transformations. To address this, we propose TanGO, a training-free framework that enables adaptive per-token steering in the tangent space of generative dynamics. To realize this selective control, we formulate a one-step optimal control rule and determine the strength of each token's control signal using a von Mises-Fisher inspired directional discrepancy derived from the source and target velocity fields. Experiments show that TanGO substantially reduces structural artifacts and achieves state-of-the-art performance, outperforming existing 3D editing baselines. The code is publicly available at https://github.com/siw00-lim/TanGO.