📊 Showing 5 results | 📏 Metric: SI-SDRi
| Rank | Model | Paper | SI-SDRi | Date | Code |
|---|---|---|---|---|---|
| 1 | SepReformer-L + DM | Separate and Reconstruct: Asymmetric Encoder-Decoder for Speech Separation | 18.40 | 2024-06-10 | 📦 dmlguq456/SepReformer |
| 2 | MossFormer (L) + DM | MossFormer: Pushing the Performance Limit of Monaural Speech Separation using Gated Single-Head Transformer with Convolution-Augmented Joint Self-Attentions | 17.30 | 2023-02-23 | 📦 modelscope/ClearerVoice-Studio 📦 alibabasglab/mossformer |
| 3 | TDANet Large | An efficient encoder-decoder architecture with top-down attention for speech separation | 15.20 | 2022-09-30 | 📦 JusperLee/TDANet |
| 4 | TDANet | An efficient encoder-decoder architecture with top-down attention for speech separation | 14.80 | 2022-09-30 | 📦 JusperLee/TDANet |
| 5 | WHYV | Wanna hear your voice? A sample is all we need! | 12.96 | 2024-10-01 | - |