Tasks SotA Datasets Papers Methods Submit About

Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable Benchmarks All SotA Datasets Papers Methods

Community

Submit Results About

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Models/Sparse HT Demucs (fine tuned)

Sparse HT Demucs (fine tuned)

Reported on 20 benchmarks across 2 tasks · 1 paper · 12 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Music10 results

Music Source SeparationonMUSDB18
SDR (avg)· uses extra data· 2022-11-15
9.2
SOTA
Hybrid Transformers for Music Source Separation arXiv:2211.08553
Music Source SeparationonMUSDB18
SDR (bass)· uses extra data· 2022-11-15
10.47
SOTA
Hybrid Transformers for Music Source Separation arXiv:2211.08553
Music Source SeparationonMUSDB18
SDR (drums)· uses extra data· 2022-11-15
10.83
SOTA
Hybrid Transformers for Music Source Separation arXiv:2211.08553
Music Source SeparationonMUSDB18-HQ
SDR (avg)· uses extra data· 2022-11-15
9.2
best: 11.99 (BS-RoFormer (L=12, OA))
SOTA
Hybrid Transformers for Music Source Separation arXiv:2211.08553
Music Source SeparationonMUSDB18-HQ
SDR (bass)· uses extra data· 2022-11-15
10.47
best: 13.32 (BS-RoFormer (L=12, OA))
SOTA
Hybrid Transformers for Music Source Separation arXiv:2211.08553
Music Source SeparationonMUSDB18-HQ
SDR (drums)· uses extra data· 2022-11-15
10.83
best: 12.91 (BS-RoFormer (L=12, OA))
SOTA
Hybrid Transformers for Music Source Separation arXiv:2211.08553
Music Source SeparationonMUSDB18
SDR (other)· uses extra data· 2022-11-15
6.41
best: 7.08 (Band-Split RNN (semi-sup.))
Hybrid Transformers for Music Source Separation arXiv:2211.08553
Music Source SeparationonMUSDB18
SDR (vocals)· uses extra data· 2022-11-15
9.37
best: 10.47 (Band-Split RNN (semi-sup.))
Hybrid Transformers for Music Source Separation arXiv:2211.08553
Music Source SeparationonMUSDB18-HQ
SDR (others)· uses extra data· 2022-11-15
6.41
best: 9.01 (BS-RoFormer (L=12, OA))
Hybrid Transformers for Music Source Separation arXiv:2211.08553
Music Source SeparationonMUSDB18-HQ
SDR (vocals)· uses extra data· 2022-11-15
9.37
best: 12.72 (BS-RoFormer (L=12, OA))
Hybrid Transformers for Music Source Separation arXiv:2211.08553

Methodology10 results

2D ClassificationonMUSDB18
SDR (avg)· uses extra data· 2022-11-15
9.2
SOTA
Hybrid Transformers for Music Source Separation arXiv:2211.08553
2D ClassificationonMUSDB18
SDR (bass)· uses extra data· 2022-11-15
10.47
SOTA
Hybrid Transformers for Music Source Separation arXiv:2211.08553
2D ClassificationonMUSDB18
SDR (drums)· uses extra data· 2022-11-15
10.83
SOTA
Hybrid Transformers for Music Source Separation arXiv:2211.08553
2D ClassificationonMUSDB18-HQ
SDR (avg)· uses extra data· 2022-11-15
9.2
best: 11.99 (BS-RoFormer (L=12, OA))
SOTA
Hybrid Transformers for Music Source Separation arXiv:2211.08553
2D ClassificationonMUSDB18-HQ
SDR (bass)· uses extra data· 2022-11-15
10.47
best: 13.32 (BS-RoFormer (L=12, OA))
SOTA
Hybrid Transformers for Music Source Separation arXiv:2211.08553
2D ClassificationonMUSDB18-HQ
SDR (drums)· uses extra data· 2022-11-15
10.83
best: 12.91 (BS-RoFormer (L=12, OA))
SOTA
Hybrid Transformers for Music Source Separation arXiv:2211.08553
2D ClassificationonMUSDB18
SDR (other)· uses extra data· 2022-11-15
6.41
best: 7.08 (Band-Split RNN (semi-sup.))
Hybrid Transformers for Music Source Separation arXiv:2211.08553
2D ClassificationonMUSDB18
SDR (vocals)· uses extra data· 2022-11-15
9.37
best: 10.47 (Band-Split RNN (semi-sup.))
Hybrid Transformers for Music Source Separation arXiv:2211.08553
2D ClassificationonMUSDB18-HQ
SDR (others)· uses extra data· 2022-11-15
6.41
best: 9.01 (BS-RoFormer (L=12, OA))
Hybrid Transformers for Music Source Separation arXiv:2211.08553
2D ClassificationonMUSDB18-HQ
SDR (vocals)· uses extra data· 2022-11-15
9.37
best: 12.72 (BS-RoFormer (L=12, OA))
Hybrid Transformers for Music Source Separation arXiv:2211.08553