Object Hallucination Evaluation on POPE GQA
91.03AccuracyRegular
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| RegularSetting=Random, Model=LLaVA1.5-13b2025.05 | 91.03 | 90.75 | |
| DoLaSetting=Random, Model=LLaVA1.5-13b2025.05 | 91.03 | 90.67 | |
| DCLASetting=Random, Model=LLaVA1.5-13b2025.05 | 91 | 90.7 | |
| DAMOSetting=Random, Model=LLaVA1.5-13b2025.05 | 90.9 | 91.09 | |
| DCLASetting=Random, Model=LLaVA-NEXT2025.05 | 90.1 | 89.52 | |
| RegularSetting=Random, Model=LLaVA-NEXT2025.05 | 90.03 | 89.34 | |
| VCDSetting=Random, Model=LLaVA1.5-13b2025.05 | 89.43 | 88.9 | |
| DAMOSetting=Random, Model=LLaVA-NEXT2025.05 | 88.63 | 87.86 | |
| DoLaSetting=Popular, Model=LLaVA1.5-13b2025.05 | 88.6 | 88.43 | |
| RegularSetting=Popular, Model=LLaVA1.5-13b2025.05 | 88.43 | 88.38 | |
| DCLASetting=Popular, Model=LLaVA1.5-13b2025.05 | 88.4 | 88.33 | |
| VCDBase Model=LLaVA-1.5-7B, Sampling Strategy=Random2026.06 | 88.1 | 89 | |
| DCLASetting=Random, Model=LLaVA1.5-7b2025.05 | 87.9 | 88.94 | |
| DoLaSetting=Random, Model=LLaVA-NEXT2025.05 | 86.97 | 85.34 | |
| GreedyBase Model=mPLUG-Owl2-7B, Sampling Strategy=Random2026.06 | 86.9 | 86.1 | |
| RegularSetting=Random, Model=LLaVA1.5-7b2025.05 | 86.87 | 88.14 | |
| DCLASetting=Random, Model=mPLUG-Owl22025.05 | 86.8 | 86.25 | |
| DAMOSetting=Popular, Model=LLaVA1.5-13b2025.05 | 86.77 | 87.55 | |
| DAMOSetting=Random, Model=mPLUG-Owl22025.05 | 86.7 | 86.23 | |
| RegularSetting=Popular, Model=LLaVA-NEXT2025.05 | 86.57 | 86.23 | |
| DCLASetting=Popular, Model=LLaVA-NEXT2025.05 | 86.57 | 86.3 | |
| VCDSetting=Popular, Model=LLaVA1.5-13b2025.05 | 86.53 | 86.28 | |
| VLIModel=Qwen3-VL2026.01 | 86.47 | 85.96 | |
| RegularSetting=Random, Model=mPLUG-Owl22025.05 | 86.1 | 85.41 | |
| DAMOSetting=Random, Model=LLaVA1.5-7b2025.05 | 86.03 | 87.52 | |
| VCDSetting=Random, Model=LLaVA1.5-7b2025.05 | 86 | 87.33 | |
| DAMOSetting=Popular, Model=LLaVA-NEXT2025.05 | 85.97 | 85.43 | |
| DoLaSetting=Random, Model=LLaVA1.5-7b2025.05 | 85.47 | 87 | |
| DoLaSetting=Popular, Model=LLaVA-NEXT2025.05 | 85.07 | 83.55 | |
| VCDModel=Qwen3-VL2026.01 | 84.84 | 84.48 | |
| DoLaSetting=Random, Model=mPLUG-Owl22025.05 | 84.77 | 83.06 | |
| CICDModel=Qwen3-VL2026.01 | 84.13 | 83.78 | |
| FADEBase Model=LLaVA-1.5-7B, Sampling Strategy=Popular2026.06 | 84.1 | 84.8 | |
| RegularSetting=Adversarial, Model=LLaVA1.5-13b2025.05 | 83.73 | 84.4 | |
| DCLASetting=Adversarial, Model=LLaVA1.5-13b2025.05 | 83.73 | 84.37 | |
| OPERAModel=Qwen3-VL2026.01 | 83.54 | 83.13 | |
| VLIModel=LLaVA-1.52026.01 | 83.49 | 83.18 | |
| DCLASetting=Adversarial, Model=LLaVA-NEXT2025.05 | 83.17 | 83.4 | |
| RegularSetting=Adversarial, Model=LLaVA-NEXT2025.05 | 82.97 | 83.27 | |
| DoLaSetting=Adversarial, Model=LLaVA1.5-13b2025.05 | 82.97 | 81.89 | |
| VTIModel=Qwen3-VL2026.01 | 82.68 | 81.97 | |
| DAMOSetting=Adversarial, Model=LLaVA-NEXT2025.05 | 82.53 | 82.49 | |
| VCDSetting=Adversarial, Model=LLaVA1.5-13b2025.05 | 82.43 | 82.81 | |
| OriginModel=Qwen3-VL2026.01 | 82.34 | 81.93 | |
| ClearSightModel=LLaVA-1.52026.01 | 81.89 | 81.64 | |
| DoLaSetting=Adversarial, Model=LLaVA-NEXT2025.05 | 81.87 | 80.71 | |
| VCDSetting=Random, Model=mPLUG-Owl22025.05 | 81.73 | 80.25 | |
| NulluModel=Qwen3-VL2026.01 | 81.48 | 80.93 | |
| DAMOSetting=Adversarial, Model=LLaVA1.5-13b2025.05 | 81.47 | 83.39 | |
| DoLaSetting=Popular, Model=mPLUG-Owl22025.05 | 81.37 | 80.29 | |
| OPERAModel=LLaVA-1.52026.01 | 81.33 | 80.97 | |
| VTIModel=LLaVA-1.52026.01 | 81.14 | 80.88 | |
| VCDSetting=Random, Model=LLaVA-NEXT2025.05 | 81.13 | 77.29 | |
| NulluModel=LLaVA-1.52026.01 | 80.43 | 79.92 | |
| DAMOSetting=Popular, Model=mPLUG-Owl22025.05 | 79.93 | 80.59 | |
| DCLASetting=Popular, Model=mPLUG-Owl22025.05 | 79.87 | 80.44 | |
| CICDModel=LLaVA-1.52026.01 | 79.54 | 80.13 | |
| RegularSetting=Popular, Model=mPLUG-Owl22025.05 | 79.37 | 80.38 | |
| ClearSightModel=Qwen3-VL2026.01 | 79.24 | 78.08 | |
| VCDSetting=Popular, Model=LLaVA-NEXT2025.05 | 79.2 | 75.53 | |
| FADEBase Model=mPLUG-Owl2-7B, Sampling Strategy=Adversarial2026.06 | 78.3 | 79 | |
| DCLASetting=Adversarial, Model=mPLUG-Owl22025.05 | 78.17 | 79.13 | |
| VCDModel=LLaVA-1.52026.01 | 78.13 | 78.54 | |
| RegularSetting=Adversarial, Model=mPLUG-Owl22025.05 | 78.07 | 79.03 | |
| DoLaSetting=Adversarial, Model=mPLUG-Owl22025.05 | 78.07 | 77.8 | |
| DAMOSetting=Adversarial, Model=mPLUG-Owl22025.05 | 78.03 | 79.13 | |
| VCDSetting=Popular, Model=mPLUG-Owl22025.05 | 77.83 | 77.12 | |
| MSCOCO→GQASettings=MSCOCO→GQA2026.03 | 77.31 | 79.57 | |
| Antidote→GQASettings=Antidote→GQA2026.03 | 77.28 | 80.12 | |
| OriginModel=LLaVA-1.52026.01 | 77.26 | 77.58 | |
| LTS-FS(Nullu)Settings=LTS-FS(Nullu)2026.03 | 77.15 | 80.63 | |
| VCDSetting=Adversarial, Model=LLaVA-NEXT2025.05 | 76.23 | 72.98 | |
| DAMOSetting=Popular, Model=LLaVA1.5-7b2025.05 | 76.17 | 80.29 | |
| VISTABase Model=InstructBLIP-7B, Sampling Strategy=Popular2026.06 | 75.8 | 78.5 | |
| FADEBase Model=InstructBLIP-7B, Sampling Strategy=Adversarial2026.06 | 75.6 | 77.6 | |
| RegularSettings=Regular2026.03 | 75.47 | 79.83 | |
| DoLaSetting=Popular, Model=LLaVA1.5-7b2025.05 | 75.3 | 79.75 | |
| DCLASetting=Popular, Model=LLaVA1.5-7b2025.05 | 75.2 | 79.68 | |
| RegularSetting=Popular, Model=LLaVA1.5-7b2025.05 | 74.5 | 79.29 | |
| VCDSetting=Adversarial, Model=mPLUG-Owl22025.05 | 74.43 | 74.53 | |
| VCDSetting=Popular, Model=LLaVA1.5-7b2025.05 | 73.73 | 78.6 | |
| DAMOSetting=Adversarial, Model=LLaVA1.5-7b2025.05 | 70.07 | 76.43 | |
| DCLASetting=Adversarial, Model=LLaVA1.5-7b2025.05 | 69.13 | 75.91 | |
| DoLaSetting=Adversarial, Model=LLaVA1.5-7b2025.05 | 68.9 | 75.77 | |
| RegularSetting=Adversarial, Model=LLaVA1.5-7b2025.05 | 68.33 | 75.5 | |
| VCDSetting=Adversarial, Model=LLaVA1.5-7b2025.05 | 68.1 | 75.15 |