二维编辑，三维验证：强化学习用于多视角一致场景编辑
Edit in 2D, Verify in 3D: Reinforcement Learning for Multi-view Consistent Scene Editing
AI总结：
本文提出RL3DEdit框架，利用强化学习优化多视角一致的3D场景编辑，通过VGGT模型生成的置信度图和姿态误差作为奖励信号，提升编辑质量与效率。
摘要：
利用2D扩散模型先验知识进行3D编辑已成为有前景的范式。然而，编辑结果的多视角一致性仍具挑战性，且配对的3D一致编辑数据极度稀缺，使监督微调（SFT）难以实施。本文观察到，尽管生成多视角一致的3D内容极具挑战性，但验证3D一致性是可行的，自然将强化学习（RL）定位为可行方案。受此启发，我们提出RL3DEdit，一个由RL优化驱动的单次通过框架，通过VGGT模型衍生的新型奖励信号，将2D编辑先验锚定在3D一致流形上。大量实验表明，RL3DEdit实现了稳定的多视角一致性，并在编辑质量和效率上优于现有最先进方法。为促进3D编辑的发展，我们将发布代码和模型。
英文摘要：
Leveraging the priors of 2D diffusion models for 3D editing has emerged as a promising paradigm. However, multi-view consistency remains challenging in edited results, and the extreme scarcity of paired 3D-consistent editing data makes supervised fine-tuning (SFT) impractical, despite its effectiveness for editing tasks. In this paper, we observe that, while generating multi-view consistent 3D content is highly challenging, verifying 3D consistency is tractable, naturally positioning reinforcement learning (RL) as a feasible solution. Motivated by this, we propose RL3DEdit, a single-pass framework driven by RL optimization with novel rewards derived from the 3D foundation model, VGGT. Specifically, we leverage VGGT's robust priors learned from massive real-world data, feed the edited images into it, and utilize the output confidence maps and pose estimation errors as reward signals, effectively anchoring the 2D editing priors onto a 3D-consistent manifold via RL. Extensive experiments demonstrate that RL3DEdit achieves stable multi-view consistency and outperforms state-of-the-art methods in editing quality with high efficiency. To promote the development of 3D editing, we will release the code and model.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
