Target Sound Extraction on AudioSet 2Mix
8.9SNRiUSE-B
Evaluation Results
| Method | Links | |
|---|---|---|
| USE-BInput Modality=text+tag+video2025.12 | 8.9 | |
| SepformerInput Modality=text+tag+video2025.12 | 8.5 | |
| USE-SInput Modality=text+tag+video2025.12 | 8.5 | |
| BSRNNInput Modality=text+tag+video2025.12 | 8.4 | |
| AudiosepInput Modality=text2025.12 | 7.7 | |
| DCCRNInput Modality=text+tag+video2025.12 | 6.9 | |
| LASSInput Modality=text2025.12 | 6.8 | |
| MAEInput Modality=audio2025.12 | 5.6 | |
| USSInput Modality=audio2025.12 | 5.6 |