Embodied Navigation on MSNN
40.44-Direction Success RatePOMA-3D_spec
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| POMA-3D_specModality=PM2025.11 | 40.4 | 21.2 | |
| 3D-ViSTAModality=C-PC2025.11 | 39.9 | 20.1 | |
| FG-CLIPModality=PM2025.11 | 39.3 | 20.4 | |
| POMA-3D_llmModality=PM, Grid inputs=H x W x 3, LoRA-tuned=true2025.11 | 36.9 | 21.4 | |
| SceneVerseModality=C-PC2025.11 | 36 | 19.5 | |
| LLaVA-OV-7BModality=RGB2025.11 | 24 | 5.83 | |
| LLaVA-3DModality=RGB-D2025.11 | 22.9 | 12.3 | |
| Qwen2.5-VL-7BModality=RGB2025.11 | 21.8 | 2.87 |