Hallucination Evaluation on POPE Adversarial v1.0 (test)
88.96AccuracyResDec
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ResDecModel=Qwen2.5-VL2026.02 | 88.96 | 87.24 | |
| VISTAModel=Qwen2.5-VL2026.02 | 88.03 | 87.24 | |
| ONLYModel=Qwen2.5-VL2026.02 | 87.32 | 86.63 | |
| OPERAModel=Qwen2.5-VL2026.02 | 86.79 | 87.35 | |
| DoLaModel=Qwen2.5-VL2026.02 | 85.67 | 84.88 | |
| VCDModel=Qwen2.5-VL2026.02 | 85.45 | 85.97 | |
| ICDModel=Qwen2.5-VL2026.02 | 85.45 | 85.97 | |
| RegularModel=Qwen2.5-VL2026.02 | 84.9 | 83.35 | |
| ResDecModel=LLaVA-1.52026.02 | 84.5 | 83.69 | |
| VISTAModel=LLaVA-1.52026.02 | 82.84 | 82.81 | |
| MemVRModel=LLaVA-1.52026.02 | 82.31 | 83.97 | |
| OA-VCDModel=LLaVA-v1.5 (7B)2026.02 | 81.9 | 82 | |
| AGLAModel=LLaVA-v1.5 (7B)2026.02 | 81.2 | 81.3 | |
| AGLAModel=LLaVA-1.52026.02 | 81.13 | 81.36 | |
| OPERAModel=LLaVA-1.52026.02 | 80.88 | 80.69 | |
| VISTAModel=InstructBLIP2026.02 | 80.82 | 79.67 | |
| ResDecModel=InstructBLIP2026.02 | 80.41 | 81.87 | |
| VCDModel=LLaVA-v1.5 (7B)2026.02 | 79.4 | 79.9 | |
| ICDModel=LLaVA-1.52026.02 | 78.16 | 78.63 | |
| RegularModel=LLaVA-v1.5 (7B)2026.02 | 77.4 | 77.4 | |
| VCDModel=LLaVA-1.52026.02 | 77.31 | 79.28 | |
| AGLAModel=InstructBLIP2026.02 | 77.29 | 79.16 | |
| DoLaModel=LLaVA-1.52026.02 | 76.32 | 76.16 | |
| VCDModel=InstructBLIP2026.02 | 76.32 | 78.08 | |
| RegularModel=LLaVA-1.52026.02 | 76.03 | 76.26 | |
| OPERAModel=InstructBLIP2026.02 | 75.5 | 78.17 | |
| ICDModel=InstructBLIP2026.02 | 75.12 | 76.24 | |
| ONLYModel=InstructBLIP2026.02 | 74.83 | 78 | |
| DoLaModel=InstructBLIP2026.02 | 74.67 | 76.68 | |
| ONLYModel=LLaVA-1.52026.02 | 72.44 | 77.49 | |
| RegularModel=InstructBLIP2026.02 | 72.37 | 75.42 |