AffordSim：一种可扩展的数据生成器和基准，用于面向 affordance 的机器人操作
AffordSim: A Scalable Data Generator and Benchmark for Affordance-Aware Robotic Manipulation
AI总结：
AffordSim 通过整合开放词汇 3D affordance 预测，解决了机器人操作中接触信息获取的挑战，实现了高成功率的轨迹生成和跨仿真到现实的迁移。
摘要：
许多日常机器人操作技能依赖于 affordance，其成功取决于机器人是否接触后续动作所需的功能物体区域。当前的仿真数据生成器通过通用抓取估计器或手动物体接触注释获取接触信息，但通用估计器无法提供任务语义，而手动注释需要为每个新物体和任务重新编写。为解决这些挑战，我们引入了 AffordSim，一种可扩展的数据生成器和基准，它将开放词汇 3D affordance 预测整合到基于仿真的轨迹生成中。给定一个自然语言任务描述，AffordSim 生成与任务相关的场景，发出 affordance 查询，将其与物体表面结合，采样区域条件抓取，并通过运动规划选择可执行的候选。它进一步随机化物体姿态、纹理、光照、图像噪声和跨视角背景以促进仿真到现实的迁移。我们实例化 AffordSim 为一个涵盖多样操作技能、五种机器人形态和 500 多个刚体和关节机械臂的 50 任务基准。AffordSim 在 affordance 关键任务上实现了 93% 的轨迹收集成功率，在硬复合任务上实现了 89% 的成功率。在 AffordSim 数据上训练的视觉-语言-动作策略能够零样本转移到真实的 Franka FR3，达到 24% 的平均成功率。
英文摘要：
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.
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。
