GeoRemover：移除物體及其因果視覺偽影

GeoRemover: Removing Objects and Their Causal Visual Artifacts

September 23, 2025

作者: Zixin Zhu, Haoxiang Li, Xuelu Feng, He Wu, Chunming Qiao, Junsong Yuan

cs.AI

摘要

迈向智能图像编辑，目标物体的移除应同时消除该物体及其因果视觉伪影，如阴影与反射。然而，现有基于图像外观的方法要么严格遵循掩码对齐训练，未能去除那些未明确掩码的因果效应；要么采用宽松的掩码对齐策略，缺乏可控性，可能无意中过度擦除其他物体。我们认识到，这些局限源于忽视了物体几何存在与其视觉效应之间的因果关系。为克服这一局限，我们提出了一种几何感知的两阶段框架，将物体移除解耦为（1）几何移除与（2）外观渲染。在第一阶段，我们利用严格掩码对齐的监督直接从几何（如深度）中移除物体，实现具有强几何约束的结构感知编辑。在第二阶段，我们基于更新后的几何条件渲染出逼真的RGB图像，其中因果视觉效应作为修改后三维几何的隐含结果被考虑。为引导几何移除阶段的学习，我们引入了一种基于正负样本对的偏好驱动目标，鼓励模型在移除物体及其因果视觉伪影的同时，避免新的结构插入。大量实验表明，我们的方法在两个流行基准测试中，在移除物体及其相关伪影方面达到了最先进的性能。代码可在https://github.com/buxiangzhiren/GeoRemover获取。

English

Towards intelligent image editing, object removal should eliminate both the target object and its causal visual artifacts, such as shadows and reflections. However, existing image appearance-based methods either follow strictly mask-aligned training and fail to remove these causal effects which are not explicitly masked, or adopt loosely mask-aligned strategies that lack controllability and may unintentionally over-erase other objects. We identify that these limitations stem from ignoring the causal relationship between an object's geometry presence and its visual effects. To address this limitation, we propose a geometry-aware two-stage framework that decouples object removal into (1) geometry removal and (2) appearance rendering. In the first stage, we remove the object directly from the geometry (e.g., depth) using strictly mask-aligned supervision, enabling structure-aware editing with strong geometric constraints. In the second stage, we render a photorealistic RGB image conditioned on the updated geometry, where causal visual effects are considered implicitly as a result of the modified 3D geometry. To guide learning in the geometry removal stage, we introduce a preference-driven objective based on positive and negative sample pairs, encouraging the model to remove objects as well as their causal visual artifacts while avoiding new structural insertions. Extensive experiments demonstrate that our method achieves state-of-the-art performance in removing both objects and their associated artifacts on two popular benchmarks. The code is available at https://github.com/buxiangzhiren/GeoRemover.