List your company

Hallucination Evaluation on A-OKVQA (POPE Accuracy Metrics)

93.76Accuracy (Random)OPERA
87.426489.070790.71592.3593Mar 17, 2026

Evaluation results

MethodLinks
OPERABackbone=Qwen3-VL2026.0393.7689.583.86
WoodpeckerBackbone=InternVL3.52026.0393.7389.884
RITUALBackbone=InternVL3.52026.0393.7189.7583.9
KestrelBackbone=InternVL3.52026.0393.5791.887.13
OPERABackbone=InternVL3.52026.0393.5589.7983.81
VCDBackbone=Qwen3-VL2026.0393.5387.8682
KestrelBackbone=Qwen3-VL2026.0393.4691.7386.76
DeGFBackbone=InternVL3.52026.0393.3989.6884.11
WoodpeckerBackbone=Qwen3-VL2026.0393.2388.983.33
DeGFBackbone=Qwen3-VL2026.0392.9687.782.61
Qwen3-VL agentBackbone=Qwen3-VL2026.0392.8785.278.03
VCDBackbone=InternVL3.52026.0392.8789.7383.73
BaselineBackbone=InternVL3.52026.0392.6787.8381.53
BaselineBackbone=Qwen3-VL2026.0392.3686.6781.87
RITUALBackbone=Qwen3-VL2026.0387.6783.577.76