Speech Enhancement on Deep Noise Suppression (DNS) Challenge

Metric: SI-SDR-WB (higher is better)

LeaderboardDataset

Loading chart...

Results

Hide extra data

Sort:

#	Model↕	SI-SDR-WB▼	Extra Data	Paper	Date↕	Code
1	TF-Locoformer (M)	23.3	No	TF-Locoformer: Transformer with Local Modeling b...	2024-08-06	Code
2	ZipEnhancer (M)	22.22	No	-	-	-
3	BSRNN-S + MRSD	21.4	No	High Fidelity Speech Enhancement with Band-split...	2022-12-01	Code
4	BSRNN-S + MGD	21.4	No	High Fidelity Speech Enhancement with Band-split...	2022-12-01	Code
5	BSRNN-S	21.3	No	High Fidelity Speech Enhancement with Band-split...	2022-12-01	Code
6	MambAttention	21.234	No	MambAttention: Mamba with Multi-Head Attention f...	2025-07-01	Code
7	ZipEnhancer (S)	21.15	No	-	-	-
8	BSRNN-16k	21.1	No	High Fidelity Speech Enhancement with Band-split...	2022-12-01	Code
9	MP-SENet	21.03	No	Explicit Estimation of Magnitude and Phase Spect...	2023-08-17	Code
10	xLSTM-SENet	20.854	No	MambAttention: Mamba with Multi-Head Attention f...	2025-07-01	Code
11	MFNET	20.31	No	A Mask Free Neural Network for Monaural Speech E...	2023-06-07	Code
12	Sudo rm -rf (U=32)	19.7	No	RemixIT: Continual self-training of speech enhan...	2022-02-17	Code
13	Sudo rm-rf (U=8)	18.6	No	Continual self-training with bootstrapped remixi...	2021-10-19	Code
14	RemixIT (w Sudo U=32)	18	Yes	Continual self-training with bootstrapped remixi...	2021-10-19	Code
15	FullSubNet	17.29	No	FullSubNet: A Full-Band and Sub-Band Fusion Mode...	2020-10-29	Code
16	FullSubNet+	16.81	No	FullSubNet+: Channel Attention FullSubNet with C...	2022-03-23	Code
17	DTLN	16.34	No	-	-	Code
18	Non-Real-Time MultiScale+	16.22	No	Phase-aware Single-stage Speech Denoising and De...	2020-06-01	Code
19	RemixIT (w Sudo U=32)	16	Yes	RemixIT: Continual self-training of speech enhan...	2022-02-17	Code
20	Noisy	9.1	No	-	-	-

#1TF-Locoformer (M)SOTA
23.3
SI-SDR-WB· 2024-08-06
TF-Locoformer: Transformer with Local Modeling by Convolution for Speech Separation and Enhancement Code
#2ZipEnhancer (M)
22.22
SI-SDR-WB
No paper
#3BSRNN-S + MRSDSOTA
21.4
SI-SDR-WB· 2022-12-01
High Fidelity Speech Enhancement with Band-split RNN Code
#4BSRNN-S + MGD
21.4
SI-SDR-WB· 2022-12-01
High Fidelity Speech Enhancement with Band-split RNN Code
#5BSRNN-S
21.3
SI-SDR-WB· 2022-12-01
High Fidelity Speech Enhancement with Band-split RNN Code
#6MambAttention
21.234
SI-SDR-WB· 2025-07-01
MambAttention: Mamba with Multi-Head Attention for Generalizable Single-Channel Speech Enhancement Code
#7ZipEnhancer (S)
21.15
SI-SDR-WB
No paper
#8BSRNN-16k
21.1
SI-SDR-WB· 2022-12-01
High Fidelity Speech Enhancement with Band-split RNN Code
#9MP-SENet
21.03
SI-SDR-WB· 2023-08-17
Explicit Estimation of Magnitude and Phase Spectra in Parallel for High-Quality Speech Enhancement Code
#10xLSTM-SENet
20.854
SI-SDR-WB· 2025-07-01
MambAttention: Mamba with Multi-Head Attention for Generalizable Single-Channel Speech Enhancement Code
#11MFNET
20.31
SI-SDR-WB· 2023-06-07
A Mask Free Neural Network for Monaural Speech Enhancement Code
#12Sudo rm -rf (U=32)SOTA
19.7
SI-SDR-WB· 2022-02-17
RemixIT: Continual self-training of speech enhancement models via bootstrapped remixing Code
#13Sudo rm-rf (U=8)SOTA
18.6
SI-SDR-WB· 2021-10-19
Continual self-training with bootstrapped remixing for speech enhancement Code
#14RemixIT (w Sudo U=32)
18
SI-SDR-WB· Extra Data· 2021-10-19
Continual self-training with bootstrapped remixing for speech enhancement Code
#15FullSubNetSOTA
17.29
SI-SDR-WB· 2020-10-29
FullSubNet: A Full-Band and Sub-Band Fusion Model for Real-Time Single-Channel Speech Enhancement Code
#16FullSubNet+
16.81
SI-SDR-WB· 2022-03-23
FullSubNet+: Channel Attention FullSubNet with Complex Spectrograms for Speech Enhancement Code
#17DTLN
16.34
SI-SDR-WB
No paperCode
#18Non-Real-Time MultiScale+SOTA
16.22
SI-SDR-WB· 2020-06-01
Phase-aware Single-stage Speech Denoising and Dereverberation with U-Net Code
#19RemixIT (w Sudo U=32)
16
SI-SDR-WB· Extra Data· 2022-02-17
RemixIT: Continual self-training of speech enhancement models via bootstrapped remixing Code
#20Noisy
9.1
SI-SDR-WB
No paper