Vision-based Robot Policy Learning on DeepMind Control
89.1Stand PerformanceORCA
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| ORCABackbone=SD 1.52025.10 | 89.1 | 76.9 | 87.6 | 50 | 68 | 74.3 | |
| TADPBackbone=SD 1.52025.10 | 89 | 69.9 | 86.6 | 41.1 | 66.9 | 70.7 | |
| Text (Simple)Backbone=SD 1.52025.10 | 87.6 | 67.9 | 84.3 | 38.8 | 66.7 | 69.1 | |
| CLIPBackbone=ViT-L/162025.10 | 87.3 | 58.3 | 54.5 | 29.9 | 67.5 | 59.5 | |
| Text (Caption)Backbone=SD 1.52025.10 | 87.2 | 68.3 | 86.2 | 37.5 | 65.1 | 68.9 | |
| CoOpBackbone=SD 1.52025.10 | 87.2 | 67.8 | 87.1 | 45 | 65.9 | 70.6 | |
| VC-1Backbone=ViT-L/162025.10 | 86.1 | 54.3 | 18.3 | 40.9 | 65.7 | 53.1 | |
| SCRBackbone=SD 1.52025.10 | 85.5 | 64.3 | 81.8 | 43.4 | 66.6 | 68.3 |