ResearchTasksAudio-Visual Target Speaker ExtractionFollowBenchmarksDataset NameSOTA MethodSortMost resultsRecently updatedMost papersApplyDataset NameSOTA methodMetricTrendResultsLast UpdatedVoxCeleb-2mix impaired visual scenarios (test)MeMo (TDSE)11.27SI-SNR27Jun 11, 2026LRS2 2-mix (test)Libri2Mix GT3.16DNSMOS22Mar 23, 2026AVSEC3 (dev)SAV-GridNet17.5SISDR8Mar 12, 2026AVSEC3 (Leaderboard Evaluation Set)Lip-Face13.259SISDR3Mar 12, 2026