Embodied Visual Tracking on EVT-Bench Distracted Tracking
67.6SRVLingNav
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| VLingNav2026.01 | 67.6 | 73.5 | 5.51 | |
| TrackVLA++2026.01 | 66.5 | 68.8 | 4.71 | |
| VLingNavmode=SFT2026.01 | 66.1 | 69.7 | 4.78 | |
| NavFoM*2026.01 | 62 | 67.9 | — | |
| NavFoM2026.01 | 61.4 | 68.2 | — | |
| TrackVLA2026.01 | 57.6 | 63.2 | 5.8 | |
| Uni-NaVid2026.01 | 31.9 | 50.1 | 21.3 | |
| EVTvisual foundation model=SoM [61] with GPT-4o [36]2026.01 | 15.7 | 35.7 | 53.3 | |
| IBVSopen-vocabulary detector=GroundingDINO [32]2026.01 | 10.6 | 28.4 | 6.14 | |
| EVT2026.01 | 3.23 | 11.2 | 47.9 | |
| PoliFormeropen-vocabulary detector=GroundingDINO [32]2026.01 | 2.62 | 13.2 | 44.5 |