Multi-agent Collaborative Transport on CHAIC Indoor High Target 1.0
0.48TROracle
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Oracleground truth knowledge=true2024.11 | 0.48 | 0.37 | 0.91 | |
| LLM+BM HelperDecision-making LLM=GPT-4, Behavior Modeling=true2024.11 | 0.38 | 0.19 | 0.29 | |
| RHP2024.11 | 0.35 | 0.11 | 0.29 | |
| VLM (GPT-4o)Decision-making VLM=GPT-4o2024.11 | 0.33 | 0.06 | 0.32 | |
| w/omode=no-helper2024.11 | 0.3 | — | — | |
| Randomstrategy=random2024.11 | 0.27 | -0.05 | 0.29 |