Language-conditioned Robot Control on Language-conditioned UAV tasks nominal dynamics
100Navigation SuccessPPO-MLP
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| PPO-MLPTraining strategy=per-task2026.04 | 100 | — | — | — | — | — | |
| AeroBridge-TTATraining strategy=per-task, Test-Time Adaptation (TTA)=false2026.04 | 100 | 98 | 100 | 100 | 100 | 99.6 | |
| AeroBridge-TTATraining strategy=multi-task2026.04 | 100 | 88 | 100 | 100 | 100 | 97.6 |