HyperAIHyperAI超神经

Command Palette

Search for a command to run...

算力平台

Music Source Separation On Musdb18 Hq

评估指标

SDR (avg)
SDR (bass)
SDR (drums)
SDR (others)
SDR (vocals)

评测结果

各个模型在此基准测试上的表现结果

Paper Title
BS-RoFormer (L=12, OA)11.9913.3212.919.0112.72-
BS-RoFormer (L=6, OA)9.8011.319.497.7310.66-
SCNet-large9.699.4910.987.4410.86SCNet: Sparse Compression Network for Music Source Separation
Sparse HT Demucs (fine tuned)9.2010.4710.836.419.37Hybrid Transformers for Music Source Separation
SCNet9.008.8210.516.769.89SCNet: Sparse Compression Network for Music Source Separation
Hybrid Transformer Demucs (f.t.)9.0010.3910.086.329.20Hybrid Transformers for Music Source Separation
Band-Split RNN (semi-sup.)8.978.1610.157.0810.47Music Source Separation with Band-split RNN
TFC-TDF-UNet (v3)8.348.458.446.869.59Sound Demixing Challenge 2023 Music Demixing Track Technical Report: TFC-TDF-UNet v3
Band-Split RNN8.247.229.016.7010.01Music Source Separation with Band-split RNN
Dual-Path TFC-TDF UNet (DTTNet)8.157.557.827.0210.21Music Source Separation Based on a Lightweight Deep Learning Framework (DTTNET: DUAL-PATH TFC-TDF UNET)
Hybrid Demucs7.688.768.245.598.13Hybrid Spectrogram and Waveform Source Separation
KUIELab-MDX-Net7.477.837.205.908.97KUIELab-MDX-Net: A Two-Stream Neural Network for Music Demixing
CWS-PResUNet6.775.936.385.848.92CWS-PResUNet: Music Source Separation with Channel-wise Subband Phase-aware ResUNet
Unmix4.1884.0734.9252.6955.06Transfer Learning with Jukebox for Music Source Separation
0 of 14 row(s) selected.
Music Source Separation On Musdb18 Hq | SOTA | HyperAI超神经