List your company

Object Hallucination Evaluation on POPE GQA (Adversarial)

84.27AccuracySIRA
-2.66120819.90747142.4761565.044829Dec 29, 2025Jan 20, 2026Feb 12, 2026Mar 7, 2026Mar 29, 2026Apr 21, 2026May 14, 2026

Evaluation results

MethodLinks
SIRABackbone=Qwen2.5-VL2026.0584.27—85.18
ICTBackbone=Qwen2.5-VL2026.0584.1—83.53
MaskCDBackbone=Qwen2.5-VL2026.0583.25—84.3
DoLaBackbone=Qwen2.5-VL2026.0582—81.51
OPERABackbone=Qwen2.5-VL2026.0580.24—80.64
VCDBackbone=Qwen2.5-VL2026.0580.01—80.75
SIRABackbone=LLaVA-v1.52026.0579.6—81.17
SIDBackbone=Qwen2.5-VL2026.0579.58—79.88
BaseBackbone=Qwen2.5-VL2026.0578.77—78.56
ICTBackbone=LLaVA-v1.52026.0577.4—80.11
MaskCDBackbone=LLaVA-v1.52026.0577.1—79.3
SIDBackbone=LLaVA-v1.52026.0576.62—78.99
VCDBackbone=LLaVA-v1.52026.0576.09—78.78
BaseBackbone=LLaVA-v1.52026.0575.08—76.06
OPERABackbone=LLaVA-v1.52026.0575—78.71
DoLaBackbone=LLaVA-v1.52026.0568.73—74.78
CoFi-DecBackbone=Qwen-VL2025.120.81690.84960.8167
DeGFBackbone=Qwen-VL2025.120.81130.84180.8057
RITUALBackbone=Qwen-VL2025.120.81070.83290.8041
M3IDBackbone=Qwen-VL2025.120.81030.82930.8094
VCDBackbone=Qwen-VL2025.120.80870.81070.808
RegularBackbone=Qwen-VL2025.120.79030.80430.7854
CoFi-DecBackbone=LLaVA-1.52025.120.75210.68340.7976
DeGFBackbone=LLaVA-1.52025.120.74070.67420.7822
CoFi-DecBackbone=InstructBLIP2025.120.73250.69680.7687
DeGFBackbone=InstructBLIP2025.120.72450.68520.7532
M3IDBackbone=InstructBLIP2025.120.71170.65790.7536
VCDBackbone=InstructBLIP2025.120.710.65750.7514
RITUALBackbone=InstructBLIP2025.120.70170.64760.7478
VCDBackbone=LLaVA-1.52025.120.68830.62260.7543
M3IDBackbone=LLaVA-1.52025.120.68670.62160.7528
RegularBackbone=LLaVA-1.52025.120.6860.62430.7484
RegularBackbone=InstructBLIP2025.120.6860.63940.731
RITUALBackbone=LLaVA-1.52025.120.68230.61750.751