arXiv cs.AI· Bohai Gu, Taiyi Wu, Dazhao Du, Jian Liu, Shuai Yang, Xiaotong Zhao, Alan Zhao, Song Guo·· 4 小时前AI 评分36
Place-it-R1：用环境感知的 MLLM 推理实现物理合理的视频物体插入
Place-it-R1: Unlocking Environment-aware Reasoning Potential of MLLM for Video Object Insertion
AI 导读
arXiv 论文提出 Place-it-R1，一个由环境感知 MLLM 推理驱动的端到端视频物体插入框架。该方法用 MLLM 分析目标环境并推断物体与场景的物理交互，将推理结果转化为语义引导和空间引导两种信号指导视频扩散模型，并引入 Spatial Direct Preference Optimization，通过区域感知的偏好目标将物理违例惩罚定位到插入物体区域。
正文 · 原文
Abstract:Video object insertion is fundamental to video editing, yet existing diffusion methods often produce visually plausible but physically inconsistent results. We present Place-it-R1, an end-to-end framework for physically plausible video object insertion driven by environment-aware MLLM reasoning. Rather than treating reasoning as a generic text prompt, Place-it-R1 uses the MLLM to analyze the target environment, infer object-scene interactions, and determine where an insertion is physically valid. The resulting reasoning is translated into two complementary forms of guidance for video diffusion: semantic guidance that describes the intended physical interaction and spatial guidance that provides a valid insertion region in each frame. To further align generation with local physical realism, we introduce Spatial Direct Preference Optimization, which leverages an MLLM to rank generated candidates, and introduces a region-aware preference objective that explicitly localizes physical-violation penalties to the inserted-object region. Place-it-R1 further offers flexible and standard modes to trade off environment adaptation and scene preservation. Extensive experiments show that Place-it-R1 produces more physically coherent and visually natural insertions than state-of-the-art methods and achieves competitive results against commercial systems.
| Comments: | this https URL |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2603.06140 [cs.CV] |
| (or arXiv:2603.06140v4 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2603.06140 arXiv-issued DOI via DataCite |
Submission history
From: Bohai Gu [view email]
[v1]
Fri, 6 Mar 2026 10:48:27 UTC (11,238 KB)
[v2]
Wed, 5 Aug 2026 03:29:38 UTC (2,854 KB)
[v3]
Thu, 22 Oct 2026 09:36:04 UTC (3,791 KB)
[v4]
Fri, 30 Oct 2026 06:28:18 UTC (3,790 KB)
来源：arXiv cs.AI · arxiv.org