Affostruction：基于生成式重建的三维可供性视觉定位
Affostruction: 3D Affordance Grounding with Generative Reconstruction
把「从部分 RGBD 观测重建完整几何」与「在完整形状上做可供性视觉定位」串成一条生成式流水线，其中稀疏体素融合实现常数复杂度的多视角特征聚合、flow-based 建模可供性分布的多义性、以及用预测可供性反向指导主动视角选择，这三件组件都可直接迁移到需要补全不可见区域的三维感知任务。
原文依据 reconstructs complete object geometry from partial RGBD observations and grounds affordances on the full shape including unobserved regions
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论本文研究从物体的 RGBD 图像进行可供性视觉定位的问题，其目标是定位与描述物体上某一动作的文本查询相对应的表面区域。现有方法仅在可见表面上预测可供性区域，而我们提出 Affostruction——一个生成式框架，它从部分 RGBD 观测中重建完整的物体几何，并在包含未观测区域在内的完整形状上进行可供性视觉定位。我们的方法引入了多视角特征的稀疏体素融合，以实现常数复杂度的生成式重建；提出了一种基于流的建模方式，用以刻画可供性分布固有的多义性；并设计了一种由预测可供性引导的主动视角选择策略。Affostruction 在具有挑战性的基准上大幅超越现有方法，在可供性视觉定位上取得 19.1 aIoU，在三维重建上取得 32.67 IoU。
英文原文摘要 915 字符
This paper addresses the problem of affordance grounding from RGBD images of an object, which aims to localize surface regions corresponding to a text query that describes an action on the object. While existing methods predict affordance regions only on visible surfaces, we propose Affostruction, a generative framework that reconstructs complete object geometry from partial RGBD observations and grounds affordances on the full shape including unobserved regions. Our approach introduces sparse voxel fusion of multi-view features for constant-complexity generative reconstruction, a flow-based formulation that captures the inherent ambiguity of affordance distributions, and an active view selection strategy guided by predicted affordances. Affostruction outperforms existing methods by large margins on challenging benchmarks, achieving 19.1 aIoU on affordance grounding and 32.67 IoU for 3D reconstruction.