List your company

Object Hallucination Evaluation on GQA Popular split

86.07AccuracyMESA
71.364475.18227982.8178Apr 9, 2026Apr 22, 2026May 6, 2026May 19, 2026Jun 2, 2026Jun 15, 2026Jun 29, 2026

Evaluation results

MethodLinks
MESAModel=Qwen-VL2026.0486.0784.89
VAFModel=Qwen-VL2026.0485.4381.35
ICTModel=Qwen-VL2026.0485.3184.15
CAIBackbone=Qwen-VL2026.0684.9784.95
VCDBackbone=Qwen-VL2026.0684.3384.82
VTIModel=Qwen-VL2026.0483.6781.31
CAIBackbone=LLaVA-1.52026.0683.4384.39
MESAModel=LLaVA-v1.52026.0482.9383.91
ICTModel=LLaVA-v1.52026.0482.8683.7
ONLYBackbone=Qwen-VL2026.0682.882.72
PAIBackbone=Qwen-VL2026.0682.6781.89
VCDModel=Qwen-VL2026.0481.8382.23
VTIModel=LLaVA-v1.52026.0481.7382.39
CAIBackbone=InstructBLIP2026.0681.282.05
VAFModel=LLaVA-v1.52026.0481.1682
ICDModel=Qwen-VL2026.0480.877.89
RegularBackbone=Qwen-VL2026.0680.7780.15
VCDModel=LLaVA-v1.52026.0480.7382.24
ICDModel=LLaVA-v1.52026.0478.5678.79
NulluModel=LLaVA-v1.52026.0477.6380.13
VCDBackbone=InstructBLIP2026.0677.5779.4
ONLYBackbone=LLaVA-1.52026.0675.9379.62
ONLYBackbone=InstructBLIP2026.0674.878.41
VanillaModel=Qwen-VL2026.0474.7666.66
NulluModel=Qwen-VL2026.0474.7166.54
VCDBackbone=LLaVA-1.52026.0674.3778.97
PAIBackbone=InstructBLIP2026.0673.7777.34
VanillaModel=LLaVA-v1.52026.0473.377.93
PAIBackbone=LLaVA-1.52026.0672.7377.6
RegularBackbone=InstructBLIP2026.0672.2775.97
RegularBackbone=LLaVA-1.52026.0671.9376.88