ResearchDatasetsVoxCeleb+DCASEFollowBenchmarksTask NameDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyTask NameDataset NameSOTA ResultTrendResultsSpeech SeparationVoxCeleb2+DCASE 3 speakers+noise (test)8.3SISDR4Speech SeparationVoxCeleb2+DCASE 2 speakers+noise (test)7.92SISDR4