ResearchBenchmarksAudio-Visual Question Answering on MUSIC-AVQA (Modality Breakdown)Audio-Visual Question Answering on MUSIC-AVQA (Modality Breakdown)Follow79.44Accuracy (Audio)Crab+70.568872.871975.17577.4781Mar 4, 2026Evaluation resultsMethodMethodLinksAccuracy (Audio)Accuracy (Visual)Accuracy (Audio-Visual)Overall AccuracyCrab+2026.0379.4492.4576.281.09QA-TIGER2026.0378.5885.1473.7477.62ST-AVQA2026.0373.8774.469.5371.59SaSR-Net2026.0373.5673.2874.6674.21PSTP-Net2026.0370.9177.2672.5773.52