Hallucination Benchmark on POPE-P
87.3AccuracyV-STAR
Evaluation Results
| Method | Links | |
|---|---|---|
| V-STARData Size=40k2026.04 | 87.3 | |
| ThinkLite-VLData Size=11k2026.04 | 86.7 | |
| Vision-R1Data Size=210k2026.04 | 85.2 | |
| VL-CogitoData Size=80k2026.04 | 85 | |
| R1-OnevisionData Size=155k2026.04 | 84 | |
| Qwen2.5VL2026.04 | 83.2 | |
| OpenVLThinkerData Size=59.2k2026.04 | 82.5 | |
| VL-RethinkerData Size=39k2026.04 | 81.8 |