HandsOn: A Benchmark for Dexterous Manipulation with 66 Tasks and 18,000 Human D
Overview
A research team from Tsinghua University, in collaboration with other institutions, has developed the HandsOn benchmark, a comprehensive evaluation framework for dexterous robotic hands. The benchmark includes 66 tasks and 18,000 human-teleoperation demonstrations, designed to assess both execution and generalization capabilities. It highlights the challenges of dexterous manipulation, including contact management and task transfer, and provides a structured approach to evaluate the performance of different strategies.
From Chinese robotics — Dataset and collection discovery
Event evidence and corrections
0 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
Reported quantity · task count: 66 other · Basis not reported · Differing source assertions
“HandsOn 包含 66 项人工验证任务和 18,000 条人类仿真遥操作示范，围绕工具使用、精细接触、双手协作和任务迁移展开评测。”
Exact source · revision 2Source owner not reported
Reported quantity · simulation demonstration count: 18000 other · Basis not reported
“HandsOn 包含 66 项人工验证任务和 18,000 条人类仿真遥操作示范，围绕工具使用、精细接触、双手协作和任务迁移展开评测。”
Exact source · revision 2Source owner not reported
Reported quantity · Reported success rate: 29.1 percent · Basis not reported · Differing source assertions
“结果显示，在提供训练示范的 Core-50 任务中，表现最好的受训多任务策略成功率为 29.1%。”
Exact source · revision 2Source owner not reported
Reported quantity · Reported success rate: 1 percent · Basis not reported · Differing source assertions
“面对没有目标任务训练示范的 Transfer-16，8 种受训多任务策略的任务平均成功率均低于 1%。”
Exact source · revision 2Source owner not reported
Reported quantity · Reported success rate: 5.2 percent · Basis not reported · Differing source assertions
“此外，GPT-6 未经过基准数据微调，通过 move_eef 接口控制手掌与手指，在 660 次测试中，Core-50 和 Transfer-16 的成功率分别为 5.2% 和 1.9%。”
Exact source · revision 2Source owner not reported
Reported quantity · Reported success rate: 1.9 percent · Basis not reported · Differing source assertions
“此外，GPT-6 未经过基准数据微调，通过 move_eef 接口控制手掌与手指，在 660 次测试中，Core-50 和 Transfer-16 的成功率分别为 5.2% 和 1.9%。”
Exact source · revision 2Source owner not reported
Reported quantity · test count: 660 other · Basis not reported
“此外，GPT-6 未经过基准数据微调，通过 move_eef 接口控制手掌与手指，在 660 次测试中，Core-50 和 Transfer-16 的成功率分别为 5.2% 和 1.9%。”
Exact source · revision 2Source owner not reported
Reported quantity · task count: 16 other · Basis not reported · Differing source assertions
“HandsOn 包含 18,000 条精选仿真遥操作示范，任务分为 Core-50 与 Transfer-16。”
Exact source · revision 2Source owner not reported
Artifact availability · paper: available
“论文 arXiv：https://arxiv.org/abs/2610.26765 代码 GitHub：https://github.com/TsinghuaC3I/HandsOn 数据 Hugging Face：https://huggingface.co/datasets/Cyan666/HandsOn-Datasets 为什么灵巧手需要一套新的评测基准？”
Exact source · revision 2Source owner not reported
Artifact availability · code: available
“论文 arXiv：https://arxiv.org/abs/2610.26765 代码 GitHub：https://github.com/TsinghuaC3I/HandsOn 数据 Hugging Face：https://huggingface.co/datasets/Cyan666/HandsOn-Datasets 为什么灵巧手需要一套新的评测基准？”
Exact source · revision 2Source owner not reported
Artifact availability · dataset: available
“论文 arXiv：https://arxiv.org/abs/2610.26765 代码 GitHub：https://github.com/TsinghuaC3I/HandsOn 数据 Hugging Face：https://huggingface.co/datasets/Cyan666/HandsOn-Datasets 为什么灵巧手需要一套新的评测基准？”
Exact source · revision 2Source owner not reported
Evidence dependency changes (3)
- 2026-10-30T19:07:55.894Z · evidence updated · source revision 2. Evidence extraction was updated; current source attributions are shown above.
- 2026-10-30T18:28:18.093Z · evidence updated · source revision 2. Evidence extraction was updated; current source attributions are shown above.
- 2026-10-30T17:37:37.351Z · evidence updated · source revision 2. Evidence extraction was updated; current source attributions are shown above.
Report timeline
Follow attributed reports and material updates.
- Chinese robotics — Dataset and collection discoverySignalHandsOn: A Benchmark for Dexterous Manipulation with 66 Tasks and 18,000 Human Demonstrations
A research team from Tsinghua University, in collaboration with other institutions, has developed the HandsOn benchmark, a comprehensive evaluation framework for dexterous robotic hands. The benchmark includes 66 tasks and 18,000 human-teleoperation demonstrations, designed to assess both execution and generalization capabilities. It highlights the challenges of dexterous manipulation, including contact management and task transfer, and provides a structured approach to evaluate the performance of different strategies.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).