arXiv cs.AI· Mingyang Li, Haofan Xu, Haowen Sun, Xinzhe Chen, Sihua Ren, Liqi Huang, Xinyang Sui, Chenyang Miao, Jiawei Ye, Qiongjie Cui, Xingyu Chen, Zeyang Liu, Xuguang Lan·· 14 小时前AI 评分32
AffordSim：面向可供性感知机器人操作的可扩展数据生成器与基准
AffordSim: A Scalable Data Generator and Benchmark for Affordance-Aware Robotic Manipulation
AI 导读
AffordSim 将开放词汇 3D 可供性预测集成进仿真轨迹生成，按自然语言任务合成场景、生成可供性查询并采样区域条件抓取，再用运动规划筛选可执行候选。其基准覆盖 50 项任务、5 种机器人本体和 500+ 刚体与铰接物体，在可供性关键任务上达到人工接触标注 93% 的轨迹采集成功率，困难复合任务为 89%。
正文 · 原文
Authors:Mingyang Li, Haofan Xu, Haowen Sun, Xinzhe Chen, Sihua Ren, Liqi Huang, Xinyang Sui, Chenyang Miao, Jiawei Ye, Qiongjie Cui, Xingyu Chen, Zeyang Liu, Xuguang Lan
Abstract:Many everyday robot manipulation skills are affordance-dependent, with success determined by whether the robot contacts the functional object region required by the subsequent action. Current simulation data generators obtain contacts from generic grasp estimators or per-object manual contact annotations, but generic estimators rank stable grasps without task semantics and often select contacts that are misaligned with the downstream action, while manual contact annotations must be rewritten for each new object and task. To solve these challenges, we introduce AffordSim, a scalable data generator and benchmark that integrates open-vocabulary 3D affordance prediction into simulation-based trajectory generation. Given a natural-language task description, AffordSim synthesizes a task-relevant scene, emits affordance queries, grounds them on object surfaces, samples region-conditioned grasps, and selects executable candidates with motion planning. It further randomizes object pose, texture, lighting, image noise, and cross-viewpoint backgrounds for sim-to-real transfer. We instantiate AffordSim as a 50-task benchmark across diverse manipulation skills, five robot embodiments, and 500+ rigid and articulated objects. AffordSim achieves 93% of the trajectory collection success rate of manual contact annotations on affordance-critical tasks and 89% on hard composite tasks. Vision-language-action policies trained on AffordSim data transfer zero-shot to a real Franka FR3, reaching 24% average success.
| Subjects: | Robotics (cs.RO); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2604.11674 [cs.RO] |
| (or arXiv:2604.11674v3 [cs.RO] for this version) | |
| https://doi.org/10.48550/arXiv.2604.11674 arXiv-issued DOI via DataCite |
Submission history
From: Mingyang Li [view email]
[v1]
Mon, 13 Apr 2026 16:21:44 UTC (37,135 KB)
[v2]
Mon, 11 May 2026 07:45:28 UTC (20,068 KB)
[v3]
Mon, 26 Oct 2026 15:05:43 UTC (19,057 KB)
来源：arXiv cs.AI · arxiv.org