Ensemble of UVR-MDX-NET-Voc_FT, Demucs4 Vocals 2023 and two MDX23C models. Algorithm gives the highest possible quality for vocal and instrumental stems.
Quality table
| Algorithm name | Multisong dataset | Synth dataset | MDX23 Leaderboard | ||
| SDR Vocals | SDR Instrumental | SDR Vocals | SDR Instrumental | SDR Vocals | |
| Ensemble of 4 models | 10.26 | 16.52 | 12.42 | 12.12 | 11.063 |
