ResearchBenchmarksCross-dataset domain transfer (Visual to Audio) on VGGSound-AVEL 40K AVE to AVVPCross-dataset domain transfer (Visual to Audio) on VGGSound-AVEL 40K AVE to AVVPFollow56.3Segment-level F1MICU-0.69214.10428.943.696May 12, 2026Evaluation resultsMethodMethodLinksSegment-level F1MICUPretraining Scale=40KPretraining Scale=40K2026.0556.3CoDAARPretraining Scale=40KPretraining Scale=40K2026.0555.8DCIDPretraining Scale=40KPretraining Scale=40K2026.0553BaselinePretraining Scale=40KPretraining Scale=40K2026.051.5