Single-view Spatial Reasoning on OMNI3D-BENCH
28.1Numeric Estimation (Count)GPT-4o
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| GPT-4oMethod Category=Baseline Models2026.03 | 28.1 | 35.5 | 66.7 | 57.2 | 42.9 | |
| Gemini1.5-ProMethod Category=Baseline Models2026.03 | 25.2 | 28.1 | 46.2 | 37.6 | 32 | |
| Llama-3.2Method Category=Baseline Models2026.03 | 24.3 | 19.3 | 47.5 | 27.4 | 25.6 | |
| pySpatialMethod Category=Visual Programming Approaches2026.03 | 22.9 | 38.6 | 72 | 54.7 | 44.2 | |
| Claude3.5-SonnetMethod Category=Baseline Models2026.03 | 22.4 | 20.6 | 62.2 | 50.6 | 32.2 | |
| VADARMethod Category=Visual Programming Approaches2026.03 | 21.7 | 35.5 | 56 | 57.6 | 40.4 | |
| SpaceMantisMethod Category=Baseline Models2026.03 | 20 | 21.7 | 50.6 | 48.2 | 30.3 | |
| ViperGPTMethod Category=Visual Programming Approaches2026.03 | 20 | 15.4 | 56 | 42.4 | 26.7 | |
| VisProgMethod Category=Visual Programming Approaches2026.03 | 2.9 | 0.9 | 54.7 | 25.9 | 13.5 |