Event
BiDex-100 Dual-Arm Dexterous Manipulation Simulation Benchmark Released
1 reports · 1 sources
Yong Wang and other authors released BiDex-100, a simulation benchmark for dual-arm dexterous manipulation, comprising 106 tasks organized along five capability dimensions: generalization, instruction following, few-shot teaching, long-horizon, and memory. Each task is paired with out-of-domain variants that apply controlled visual and spatial changes to evaluate robustness; the benchmark supports two embodiments and provides 8,080 successfully validated teleoperation demonstrations, totaling 6.25 million frames. The authors evaluated seven recent vision-language-action models under a unified protocol, establishing baselines for capability-specific performance, robustness, and data efficiency.
Report timeline
- BiDex-100: A Bimanual Dexterous Manipulation Benchmark Covering 100+ Simulation Tasks
The authors propose BiDex-100, a simulation benchmark for bimanual dexterous manipulation comprising 106 tasks organized along five capability dimensions: generalization, instruction following, limited demonstrations, long horizon, and memory. Each task is paired with an out-of-domain variant featuring controlled visual and spatial changes for evaluating robustness. The benchmark supports two embodiments and provides 8,080 success-validated teleoperated demonstrations totaling 6.25 million frames. The authors evaluate seven recent vision-language-action models under a unified protocol, establishing baselines for capability-specific performance, robustness, and data efficiency.
arXiv cs.RO ·