Long-horizon robot manipulation on Sapien Simulation Task 5 physical simulation (test)
100Success Rate (SR)TP-S
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| TP-SArm Configuration=Single-arm2024.06 | 100 | 8.2 | 55.8 | 64 | |
| RoCoPlanning Approach=Iterative multi-agent collaboration2024.06 | 100 | 54.4 | 35.3 | 89.7 | |
| DAG-PlanRepresentation=Task graph2024.06 | 100 | 8.3 | 35.4 | 43.7 | |
| TP-DArm Configuration=Dual-arm2024.06 | 50 | 8.3 | 43.2 | 51.4 | |
| TwoStepDecomposition=Independent sub-tasks2024.06 | 50 | 11.3 | 43.2 | 54.5 |