PerAct

Reported on 5 benchmarks across 1 task · 1 paper · 1 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Robots5 results

Robot ManipulationonRLBench
Succ. Rate (18 tasks, 10 demo/task)· 2022-09-12
30
best: 60.1 (EquAct)
SOTA
Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation arXiv:2209.05451
Robot ManipulationonRLBench
Input Image Size· 2022-09-12
128
best: 256 (EquAct)
Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation arXiv:2209.05451
Robot ManipulationonRLBench
Succ. Rate (18 tasks, 100 demo/task)· 2022-09-12
42.7
best: 89.4 (EquAct)
Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation arXiv:2209.05451
Robot ManipulationonRLBench
Training Time (V100 x 8 x day)· 2022-09-12
16
Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation arXiv:2209.05451
Robot ManipulationonThe COLOSSEUM
Average decrease average across all perturbations· 2022-09-12
-17.3
best: -4.3 (SAM2Act)
Perceiver-Actor: A Multi-Task Transformer for Robotic Manipulation arXiv:2209.05451