Locomotion on Obstacles Simulated
92.8Traversing RateTeacher
Evaluation results
| Method | Links | ||
|---|---|---|---|
| TeacherPolicy Type=Privileged elevation-map-based policy2023.04 | 92.8 | 87.8 | |
| NVMCore Mechanism=Neural Volumetric Memory2023.04 | 74.1 | 61.6 | |
| NaiveCNN-RNNMemory=2-layer GRU2023.04 | 67.9 | 49.8 | |
| LocoTransformerModel=Transformer2023.04 | 65.2 | 46.7 | |
| NaiveCNN-MSEncoder=2D CNN2023.04 | 63.7 | 48.1 | |
| NaiveCNNEncoder=2D CNN2023.04 | 63.3 | 45.2 |