Open-Vocabulary Object Navigation on OVON Seen (val)
65SRHydra-Nav-IRFT (Stage 3)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Hydra-Nav-IRFT (Stage 3)Observation=RGB + Depth, Action Output=Low + High, Dual Architecture=true2026.02 | 65 | 34.4 | |
| Nav-R1Observation=RGB + Depth, Action Output=Low + High, Dual Architecture=true2026.02 | 58.4 | 26.3 | |
| Hydra-Nav-SFT (Stage 2)Observation=RGB + Depth, Action Output=Low + High, Dual Architecture=true2026.02 | 57.2 | 28.3 | |
| MTU3DObservation=RGB + Depth, Action Output=High, Dual Architecture=false2026.02 | 55 | 23.6 | |
| Nav-R2Observation=RGB, Action Output=Low, Dual Architecture=false2026.02 | 45.6 | 21 | |
| Uni-NavidObservation=RGB, Action Output=Low, Dual Architecture=false2026.02 | 41.3 | 21.1 | |
| NavFoMObservation=RGB, Action Output=Low, Dual Architecture=false2026.02 | 40.1 | 27.1 | |
| VLFMObservation=RGB + Depth, Action Output=High, Dual Architecture=false2026.02 | 35.2 | 18.6 | |
| Hydra-Nav-Base (Stage 1)Observation=RGB + Depth, Action Output=Low + High, Dual Architecture=true2026.02 | 31.9 | 25.8 |