List your company

Multimodal Understanding on POPE (test)

87.3AccuracySpatial Relation
86.36486.60786.8587.093Jul 5, 2026

Evaluation results

MethodLinks
Spatial RelationTrain (und)=Spatial Relation2026.0787.3
BaselineTrain (und)=Baseline2026.0786.7
Text recognitionTrain (und)=Text recognition2026.0786.7
CountingTrain (und)=Counting2026.0786.4