User-Centric Continual Spatial Intelligence Reasoning on UCS-Bench
91.7Trajectory & Movement (Overall)Human
Evaluation results
| Method | Links | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| HumanSize=-, # Frame=-2026.06 | 91.7 | 90.9 | 94.1 | 87.6 | 84 | 89.8 | 96.8 | 95 | 98.2 | 90 | 94.3 | 88.1 | 91 | |
| DirectMe (w/ Qwen3-VL)Size=8B, # Frame=1fps2026.06 | 54.1 | 53 | 57.2 | 47.6 | 50.2 | 46 | 58 | 55.7 | 59.8 | 46.4 | 53.7 | 43.1 | 50.5 | |
| InternVL3Size=38B, # Frame=642026.06 | 53.7 | 52.4 | 57.3 | 45.6 | 46.9 | 44.8 | 55.3 | 54 | 56.3 | 45.8 | 60.4 | 39.3 | 49.2 | |
| DirectMe (w/ InternVL3)Size=8B, # Frame=1fps2026.06 | 52.6 | 51.6 | 55.4 | 47.1 | 47.3 | 47 | 56.9 | 54.2 | 59.1 | 46 | 54.5 | 42.2 | 49.7 | |
| Qwen2.5-VLSize=32B, # Frame=642026.06 | 50.3 | 49.4 | 53 | 40.9 | 43.1 | 39.5 | 55.4 | 50.4 | 59.3 | 46 | 56.4 | 41.4 | 47.3 | |
| Qwen3-VLSize=32B, # Frame=642026.06 | 50.3 | 49 | 54.1 | 46.5 | 50.1 | 44.3 | 55.2 | 53.9 | 56.3 | 46.7 | 58.8 | 41.2 | 49 | |
| InternVL3.5Size=38B, # Frame=642026.06 | 50.2 | 49.5 | 52.2 | 46.3 | 45.9 | 46.6 | 53.9 | 53.5 | 54.2 | 45 | 56.5 | 40.1 | 48.1 | |
| Doubao-seed-1.6VSize=-, # Frame=502026.06 | 49.6 | 49.4 | 50.1 | 42.8 | 46.4 | 40.4 | 56 | 54.9 | 56.8 | 44.6 | 57 | 39 | 47.3 | |
| InternVL3Size=8B, # Frame=642026.06 | 44.9 | 44.4 | 46.5 | 39.1 | 42.2 | 37.2 | 52.2 | 48.6 | 55.1 | 42.2 | 53.5 | 37.1 | 43.8 | |
| Qwen3-VLSize=8B, # Frame=642026.06 | 44.6 | 43.1 | 48.7 | 40.1 | 43.6 | 37.9 | 50.9 | 49.2 | 52.1 | 42.9 | 52.8 | 38.4 | 44 | |
| GPT-5Size=-, # Frame=502026.06 | 44.5 | 43.3 | 48.1 | 37.8 | 40.2 | 36.3 | 55.4 | 52.9 | 57.5 | 41.4 | 58.8 | 33.5 | 43.7 | |
| InternVideo2.5Size=8B, # Frame=642026.06 | 44.2 | 43.9 | 45.2 | 39.4 | 42.4 | 37.5 | 52.7 | 49.1 | 55.5 | 42.2 | 53 | 37.4 | 43.8 | |
| Cambrian-SSize=7B, # Frame=642026.06 | 44.2 | 44 | 44.5 | 35.9 | 36.2 | 35.7 | 52.3 | 43.7 | 59 | 40.8 | 50.4 | 36.6 | 42.4 | |
| InternVL3.5Size=8B, # Frame=642026.06 | 44 | 42.3 | 48.8 | 39.8 | 40.6 | 39.2 | 52.9 | 49.3 | 55.7 | 39.2 | 47.9 | 35.2 | 42.9 | |
| Gemini3-FlashSize=-, # Frame=502026.06 | 43.8 | 43.2 | 45.5 | 38 | 38.7 | 37.6 | 52.9 | 50.2 | 54.9 | 38.9 | 56.2 | 31.1 | 42.2 | |
| DeepseekV3Size=-, # Frame=502026.06 | 42 | 41 | 44.8 | 33.2 | 35.4 | 31.6 | 37.7 | 45.9 | 31.2 | 32.3 | 36.8 | 30.3 | 35.6 | |
| Qwen2.5-VLSize=7B, # Frame=642026.06 | 41.6 | 41 | 43.1 | 40.8 | 42.8 | 39.5 | 51.3 | 47.4 | 54.3 | 42.5 | 50.1 | 39.2 | 43.4 | |
| LLaVA-VideoSize=7B, # Frame=642026.06 | 40.4 | 40.9 | 38.7 | 38.7 | 42 | 36.6 | 47.2 | 45.9 | 48.2 | 41.8 | 52.4 | 37.1 | 41.7 | |
| DispiderSize=7B, # Frame=1fps2026.06 | 40.2 | 40.6 | 38.7 | 29.1 | 30.1 | 28.4 | 51.6 | 44.9 | 53.5 | 40.2 | 38.7 | 29.4 | 36.7 | |
| SenseNova-SISize=8B, # Frame=642026.06 | 38.2 | 38.7 | 36.3 | 30.8 | 32 | 30 | 49 | 47.2 | 50.3 | 37 | 45.5 | 33.3 | 37.8 | |
| ViLaSRSize=7B, # Frame=642026.06 | 37.3 | 38.1 | 34.9 | 33.2 | 34.5 | 32.3 | 52.3 | 46 | 57.2 | 40.7 | 48.7 | 37.1 | 40.1 | |
| VideoChat-OnlineSize=4B, # Frame=1fps2026.06 | 35.7 | 36.3 | 33.7 | 30.3 | 32.2 | 29 | 51.9 | 45.7 | 56.8 | 36.4 | 43.1 | 33.5 | 37.5 | |
| VG-LLMSize=8B, # Frame=642026.06 | 34.6 | 35.7 | 31.3 | 35 | 38.3 | 32.8 | 51.9 | 45.9 | 56.6 | 35.9 | 47.3 | 30.9 | 38.3 | |
| Spatial-MLLMSize=3B, # Frame=642026.06 | 33.2 | 34.1 | 30.6 | 26.1 | 28.6 | 24.4 | 53.6 | 46 | 59.5 | 30.9 | 36.3 | 28.5 | 34.3 | |
| Flash-VStreamSize=7B, # Frame=1fps2026.06 | 26.1 | 27.3 | 22.1 | 23 | 23.7 | 22.5 | 48.8 | 44.4 | 52.2 | 23.4 | 27.2 | 21.8 | 28.5 | |
| RandomSize=-, # Frame=-2026.06 | 19.1 | 20.8 | 14.2 | 22.7 | 24.4 | 21.5 | 49.7 | 51.8 | 48.1 | 19.4 | 20.3 | 19 | 25.7 |