发表机构
Korea Advanced Institute of Science and Technology; Chung-Ang University（韩国科学技术院; 中央大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究针对3D生成模型免训练编辑存在语义伪影的问题，提出TanGO框架，通过在切线空间自适应逐令牌引导，制定最优控制规则并确定控制信号强度，减少结构伪影，性能优于现有基线。
AI 中文摘要
近期基于流匹配的3D生成模型（如VecSet）采用结构化表示，但令牌共享全局上下文，导致传统免训练编辑存在语义伪影。为此提出TanGO，一个免训练框架，能在生成动力学的切线空间中实现自适应逐令牌引导。通过制定单步最优控制规则，并利用从源和目标速度场导出的冯·米塞斯 - 费舍尔启发的方向差异确定每个令牌控制信号的强度。实验表明TanGO显著减少结构伪影并实现了优于现有3D编辑基线的性能。
英文摘要
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.
CommentsECCV 2026
