1-Phase Room Rearrangement on iTHOR challenge 2021 (test)
0.9653FS (Final Success)Expert
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Experttype=Heuristic2022.11 | 0.9653 | 0.928 | — | — | 1 | |
| ASK4HELP2022.11 | 0.9532 | 0.904 | — | — | 0.39 | |
| Naive Helperprobability=0.72022.11 | 0.9485 | 0.896 | — | — | 0.7 | |
| Model Confusionepsilon=0.32022.11 | 0.9464 | 0.899 | — | — | 0.63 | |
| Naive Helperprobability=0.62022.11 | 0.9277 | 0.876 | — | — | 0.6 | |
| Naive Helperprobability=0.52022.11 | 0.9024 | 0.836 | — | — | 0.5 | |
| Model Confusionepsilon=0.22022.11 | 0.8896 | 0.814 | — | — | 0.48 | |
| Naive Helperprobability=0.42022.11 | 0.8597 | 0.775 | — | — | 0.4 | |
| EmbCLIP2022.11 | 0.1813 | 0.069 | — | — | 0 | |
| ResNet-50 (CLIP)Backbone=ResNet-50, Pre-training=CLIP2021.11 | 0.17 | 0.08 | 0.89 | 0.88 | — | |
| EmbCLIPBackbone=ResNet-50, Pre-training=CLIP2021.11 | 0.17 | 0.08 | 0.89 | 0.88 | — | |
| RN18 + ANM ILBackbone=ResNet-18, Methodology=Imitation Learning + Active Neural SLAM2021.11 | 0.09 | 0.03 | 1.04 | 1.05 | — | |
| ResNet-50 (ImageNet)Backbone=ResNet-50, Pre-training=ImageNet2021.11 | 0.07 | 0.03 | 1.06 | 1.05 | — | |
| RN18 + ILBackbone=ResNet-18, Methodology=Imitation Learning2021.11 | 0.06 | 0.03 | 1.09 | 1.11 | — |