ResearchBenchmarksDirect-Answer on A-OKVQA 1.0 (val)For AI assistantsMarkdownContent indexDirect-Answer on A-OKVQA 1.0 (val)Follow0.683AccuracyProphet++0.290920.392710.49450.59629Mar 3, 2023Evaluation resultsMethodMethodLinksAccuracyProphet++Backbone=mPLUGBackbone=mPLUG2023.030.683ProphetBackbone=mPLUGBackbone=mPLUG2023.030.647mPLUG2023.030.591ProphetBackbone=MCANBackbone=MCAN2023.030.582PromptCapBackbone=OFABackbone=OFA2023.030.563MCAN2023.030.52GPV-22023.030.486KRISP2023.030.337ClipCap2023.030.309LXMERT2023.030.307ViLBERT2023.030.306