Speech Enhancement On Deep Noise Suppression

Metrics

PESQ-WB

SI-SDR-WB

Results

Performance results of various models on this benchmark

Model Name	PESQ-WB	SI-SDR-WB	Paper Title	Repository
BSRNN-S	3.42	21.3	High Fidelity Speech Enhancement with Band-split RNN	-
Noisy	1.58	9.1	-	-
RemixIT (w Sudo U=32)	2.34	16.0	RemixIT: Continual self-training of speech enhancement models via bootstrapped remixing	-
SN-Net	-	-	Interactive Speech and Noise Modeling for Speech Enhancement	-
Sudo rm -rf (U=32)	2.95	19.7	RemixIT: Continual self-training of speech enhancement models via bootstrapped remixing	-
ZipEnhancer (S)	3.69	21.15	-	-
aTENNuate	2.98	-	aTENNuate: Optimized Real-time Speech Enhancement with Deep SSMs on Raw Audio	-
BSRNN-S + MGD	-	21.4	High Fidelity Speech Enhancement with Band-split RNN	-
DCTCRN-S	2.77	-	Real-time Monaural Speech Enhancement With Short-time Discrete Cosine Transform	-
MFNET	3.43	20.31	A Mask Free Neural Network for Monaural Speech Enhancement	-
BSRNN-16k	3.45	21.1	High Fidelity Speech Enhancement with Band-split RNN	-
ZipEnhancer (M)	3.81	22.22	-	-
DCTCRN-P	2.82	-	Real-time Monaural Speech Enhancement With Short-time Discrete Cosine Transform	-
TF-Locoformer (M)	3.72	23.3	TF-Locoformer: Transformer with Local Modeling by Convolution for Speech Separation and Enhancement	-
DCCRN-E-Aug	-	-	DCCRN: Deep Complex Convolution Recurrent Network for Phase-Aware Speech Enhancement	-
FRCRN	3.23	-	Monaural Speech Enhancement with Complex Convolutional Block Attention Module and Joint Time Frequency Losses	-
DTLN	-	16.34	Dual-Signal Transformation LSTM Network for Real-Time Noise Suppression	-
FullSubNet+	3.218	16.81	FullSubNet+: Channel Attention FullSubNet with Complex Spectrograms for Speech Enhancement	-
DCTCRN-T	2.82	-	Real-time Monaural Speech Enhancement With Short-time Discrete Cosine Transform	-
Sudo rm-rf (U=8)	2.69	18.6	Continual self-training with bootstrapped remixing for speech enhancement	-

0 of 34 row(s) selected.