作娱乐 dataset 的 Affordance Demonstrations
机器人学
2025-07-24 v1 人机交互
摘要
Affordance -即环境或其中对象提供的行动可能性-对 operate in human 环境的机器人至关重要。现有方法在标注静态图像或形状上训练此类能力。本工作提出了用于 common household 任务 affordance 学习的新型数据集。不同于以往方法，本数据集由演示任务的视频序列组成，这些视频从第一人称和第三人称视角拍摄，包含有关任务中显现 affordance 的元数据，旨在训练识别 affordance 显现形式的感知系统。演示来自多个参与者，总计记录约七小时的人类活动。任务表现的多样性也允许研究人们为任务可能执行的准备动作，例如如何布置任务空间，这对于协作服务机器人也同样重要。
关键词
引用
@article{arxiv.2507.17401,
title = {The Wilhelm Tell Dataset of Affordance Demonstrations},
author = {Rachel Ringe and Mihai Pomarlan and Nikolaos Tsiogkas and Stefano De Giorgis and Maria Hedblom and Rainer Malaka},
journal= {arXiv preprint arXiv:2507.17401},
year = {2025}
}
备注
\c{opyright} 2025 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works