ResearchBenchmarksCross-modal classification (Visual to Audio) on VGGSound-AVEL 90K UCF to VGGCross-modal classification (Visual to Audio) on VGGSound-AVEL 90K UCF to VGGFollow73.7PrecisionMICU17.0231.73546.4561.165May 12, 2026Evaluation resultsMethodMethodLinksPrecisionMICUPretraining Scale=90KPretraining Scale=90K2026.0573.7CoDAARPretraining Scale=90KPretraining Scale=90K2026.0571.2DCIDPretraining Scale=90KPretraining Scale=90K2026.0569.9BaselinePretraining Scale=90KPretraining Scale=90K2026.0519.2