List your company

Object Hallucination Assessment on A-OKVQA POPE (Adversarial)

0.8363AccuracySIRA
0.6640760.7087880.75350.798212Jan 3, 2025Apr 3, 2025Jul 2, 2025Oct 1, 2025Dec 30, 2025Mar 30, 2026Jun 29, 2026

Evaluation results

MethodLinks
SIRABackbone=Qwen2.5-VL2026.050.8363—0.8461—
ICTBackbone=Qwen2.5-VL2026.050.836—0.8302—
MaskCDBackbone=Qwen2.5-VL2026.050.8275—0.8395—
CAIBackbone=Qwen-VL2026.060.826—0.832—
ONLYBackbone=Qwen-VL2026.060.822—0.8274—
VCDBackbone=Qwen-VL2026.060.8213—0.8292—
PAIBackbone=Qwen-VL2026.060.821—0.8158—
DoLaBackbone=Qwen2.5-VL2026.050.8183—0.8133—
VCDBackbone=Qwen2.5-VL2026.050.8127—0.8238—
CoFi-DecModel=Qwen-VL2025.120.81260.80970.8104—
RegularBackbone=Qwen-VL2026.060.8103—0.8036—
SIDBackbone=Qwen2.5-VL2026.050.8084—0.8151—
OPERABackbone=Qwen2.5-VL2026.050.8082—0.8154—
DeGFModel=Qwen-VL2025.120.80750.80370.8046—
BaseBackbone=Qwen2.5-VL2026.050.8037—0.7968—
RITUALModel=Qwen-VL2025.120.8020.79080.8058—
M3IDModel=Qwen-VL2025.120.7950.77540.8021—
VCDModel=Qwen-VL2025.120.79130.76040.803—
CMPCBase Model=LLaVA 1.52025.010.79130.76450.80140.842
CAIBackbone=LLaVA-1.52026.060.774—0.7979—
SIRABackbone=LLaVA-v1.52026.050.773—0.7943—
ICDBase Model=LLaVA 1.52025.010.7720.75470.77950.806
RegularModel=Qwen-VL2025.120.7690.75590.7748—
MaskCDBackbone=LLaVA-v1.52026.050.759—0.784—
CMPCBase Model=InstructBLIP2025.010.7570.71380.77930.858
ICTBackbone=LLaVA-v1.52026.050.7556—0.7868—
SIDBackbone=LLaVA-v1.52026.050.755—0.7794—
CAIBackbone=InstructBLIP2026.060.7523—0.7804—
VCDBackbone=LLaVA-v1.52026.050.7497—0.7773—
VCDBase Model=LLaVA 1.52025.010.74970.70010.77730.8736
ICDBase Model=InstructBLIP2025.010.74420.70240.78480.8893
VCDBase Model=InstructBLIP2025.010.74330.69460.77190.8687
BaseBackbone=LLaVA-v1.52026.050.7404—0.7515—
BaselineBase Model=LLaVA 1.52025.010.74040.72080.75150.7849
OPERABackbone=LLaVA-v1.52026.050.7382—0.7791—
CoFi-DecModel=InstructBLIP2025.120.73440.66970.7716—
VCDBackbone=InstructBLIP2026.060.7333—0.77—
DeGFModel=InstructBLIP2025.120.71870.65650.7596—
CoFi-DecModel=LLaVA-1.52025.120.7130.6810.7826—
VCDModel=InstructBLIP2025.120.710.65410.7545—
BaselineBase Model=InstructBLIP2025.010.70710.65910.75560.8583
DeGFModel=LLaVA-1.52025.120.7070.6670.7686—
ONLYBackbone=LLaVA-1.52026.060.7047—0.7586—
RITUALModel=InstructBLIP2025.120.70270.64150.7555—
M3IDModel=InstructBLIP2025.120.7010.64280.7516—
RegularModel=InstructBLIP2025.120.69870.64540.7454—
ONLYBackbone=InstructBLIP2026.060.6877—0.7495—
PAIBackbone=InstructBLIP2026.060.6867—0.7455—
M3IDModel=LLaVA-1.52025.120.6860.62220.7511—
RITUALModel=LLaVA-1.52025.120.68570.62260.7499—
RegularBackbone=InstructBLIP2026.060.6833—0.7389—
VCDBackbone=LLaVA-1.52026.060.683—0.7486—
DoLaBackbone=LLaVA-v1.52026.050.6803—0.7449—
PAIBackbone=LLaVA-1.52026.060.676—0.7423—
VCDModel=LLaVA-1.52025.120.6740.61390.7421—
RegularModel=LLaVA-1.52025.120.67230.61560.737—
RegularBackbone=LLaVA-1.52026.060.6707—0.737—