RVT-2

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, 100 demo/task)· 2024-06-12
81.4
best: 89.4 (EquAct)
SOTA
RVT-2: Learning Precise Manipulation from Few Demonstrations arXiv:2406.08545
Robot ManipulationonRLBench
Inference Speed (fps)· 2024-06-12
20.6
best: 1.4 (EquAct)
RVT-2: Learning Precise Manipulation from Few Demonstrations arXiv:2406.08545
Robot ManipulationonRLBench
Input Image Size· 2024-06-12
128
best: 256 (EquAct)
RVT-2: Learning Precise Manipulation from Few Demonstrations arXiv:2406.08545
Robot ManipulationonRLBench
Training Time (V100 x 8 x day)· 2024-06-12
0.83
best: 16 (PerAct (Evaluated in RVT))
RVT-2: Learning Precise Manipulation from Few Demonstrations arXiv:2406.08545
Robot ManipulationonGEMBench
Average Success Rate· 2024-06-12
44
best: 48 (3D-LOTUS++)
RVT-2: Learning Precise Manipulation from Few Demonstrations arXiv:2406.08545