ResearchTasksAudio Understanding RobustnessFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedMMAU Sound (SNR=-10dB)EchoDistill66.83Accuracy36May 26, 2026MMAU Speech SNR=-10dBEchoDistill62.89Accuracy18May 26, 2026MMAU Music SNR=-10dBDFL64.82Accuracy18May 26, 2026