Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware
Open paper
Benchmark
Model
Rank
Results
robot-manipulation-generalization-on-the-colosseum
ACT
–
Average decrease average across all perturbations: -61.8
robot-manipulation-on-mimicgen
ACT (Evaluated in EquiDiff)
–
Succ. Rate (12 tasks, 100 demo/task): 21.3
…
Rank counts only results with a code link.