Papers with Code · CC BY-SA 4.0

Deep Noise Suppression (DNS) Challenge

아카이브 평가 36건 · 마지막 적재/변경 2026-10-30

평가 기록

과거 수집 기록이며 최신 순위가 아닙니다. 데이터셋 버전·split·실험 조건과 지표 방향이 검증되지 않은 결과는 직접 비교할 수 없습니다. 날짜는 논문 날짜이며 최신 평가일을 뜻하지 않습니다.

표가 화면보다 넓으면 좌우로 스크롤하여 논문과 구현 링크를 확인하세요.

원본 기록 순서 · 순위 미부여
모델	원본 지표·값	논문	구현
ZipEnhancer (M)	SI-SDR-WB 22.22 PESQ-NB 4.08 PESQ-WB 3.81 STOI 98.65 Number of parameters (M) 11.34 FLOPS (G) 266.96	연결 없음 논문 날짜 미기재	연결 없음
TF-Locoformer (M)	SI-SDR-WB 23.3 PESQ-WB 3.72 STOI 98.8 Number of parameters (M) 15 FLOPS (G) 497.24	TF-Locoformer: Transformer with Local Modeling by Convolution for Speech Separation and Enhancement 2024-08-06	코드 1
ZipEnhancer (S)	SI-SDR-WB 21.15 PESQ-NB 3.99 PESQ-WB 3.69 STOI 98.32 Number of parameters (M) 2.04 FLOPS (G) 62.85	연결 없음 논문 날짜 미기재	연결 없음
MambAttention	SI-SDR-WB 21.234 PESQ-WB 3.671 Number of parameters (M) 2.33 SSNR 15.116 ESTOI 95.9	MambAttention: Mamba with Multi-Head Attention for Generalizable Single-Channel Speech Enhancement 2025-07-01	코드 1 · 코드 2
MP-SENet	SI-SDR-WB 21.03 PESQ-NB 3.92 PESQ-WB 3.62	Explicit Estimation of Magnitude and Phase Spectra in Parallel for High-Quality Speech Enhancement 2023-08-17	코드 1
xLSTM-SENet	SI-SDR-WB 20.854 PESQ-WB 3.588 Number of parameters (M) 2.20 SSNR 14.526 ESTOI 95.4	MambAttention: Mamba with Multi-Head Attention for Generalizable Single-Channel Speech Enhancement 2025-07-01	코드 1 · 코드 2
BSRNN-S + MRSD	SI-SDR-WB 21.4 PESQ-NB 3.89 PESQ-WB 3.53 STOI 98.4	High Fidelity Speech Enhancement with Band-split RNN 2022-12-01	코드 1
BSRNN-16k	SI-SDR-WB 21.1 PESQ-NB 3.87 PESQ-WB 3.45 STOI 98.3	High Fidelity Speech Enhancement with Band-split RNN 2022-12-01	코드 1
MFNET	SI-SDR-WB 20.31 PESQ-NB 3.74 PESQ-WB 3.43	A Mask Free Neural Network for Monaural Speech Enhancement 2023-06-07	코드 1 · 코드 2
BSRNN-S	SI-SDR-WB 21.3 PESQ-WB 3.42	High Fidelity Speech Enhancement with Band-split RNN 2022-12-01	코드 1
BSRNN	PESQ-NB 3.79 PESQ-WB 3.32 STOI 98	High Fidelity Speech Enhancement with Band-split RNN 2022-12-01	코드 1
CleanUNet-2	PESQ-NB 3.658 PESQ-WB 3.262	CleanUNet 2: A Hybrid Speech Denoising Model on Waveform and Spectrogram 2023-09-12	연결 없음
FRCRN	PESQ-WB 3.23	Monaural Speech Enhancement with Complex Convolutional Block Attention Module and Joint Time Frequency Losses 2021-02-03	코드 1 · 코드 2
FullSubNet+	SI-SDR-WB 16.81 PESQ-NB 3.666 PESQ-WB 3.218	FullSubNet+: Channel Attention FullSubNet with Complex Spectrograms for Speech Enhancement 2022-03-23	코드 1 · 코드 2
CleanUNet	PESQ-NB 3.551 PESQ-WB 3.146	Speech Denoising in the Waveform Domain with Self-Attention 2022-02-15	코드 1
aTENNuate	PESQ-WB 2.98	aTENNuate: Optimized Real-time Speech Enhancement with Deep SSMs on Raw Audio 2024-09-05	연결 없음
Sudo rm -rf (U=32)	SI-SDR-WB 19.7 PESQ-WB 2.95	RemixIT: Continual self-training of speech enhancement models via bootstrapped remixing 2022-02-17	코드 1 · 코드 2
DCTCRN-P	PESQ-WB 2.82	Real-time Monaural Speech Enhancement With Short-time Discrete Cosine Transform 2021-02-09	연결 없음
DCTCRN-T	PESQ-WB 2.82	Real-time Monaural Speech Enhancement With Short-time Discrete Cosine Transform 2021-02-09	연결 없음
DCCRN-E	PESQ-WB 2.79	Real-time Monaural Speech Enhancement With Short-time Discrete Cosine Transform 2021-02-09	연결 없음
PoCoNet	PESQ-WB 2.7885	PoCoNet: Better Speech Enhancement with Frequency-Positional Embeddings, Semi-Supervised Conversational Data, and Biased Loss 2020-08-11	연결 없음
FullSubNet	SI-SDR-WB 17.29 PESQ-NB 3.305 PESQ-WB 2.777	FullSubNet: A Full-Band and Sub-Band Fusion Model for Real-Time Single-Channel Speech Enhancement 2020-10-29	코드 1 · 코드 2 · 코드 3 · 코드 4 · 코드 5 · 코드 6
DCTCRN-S	PESQ-WB 2.77	Real-time Monaural Speech Enhancement With Short-time Discrete Cosine Transform 2021-02-09	연결 없음
RNN-Modulation	PESQ-WB 2.75	A Modulation-Domain Loss for Neural-Network-based Real-time Speech Enhancement 2021-02-15	코드 1
Conv-TasNet-SNR	PESQ-WB 2.73	Exploring the Best Loss Function for DNN-Based Low-latency Speech Enhancement with Temporal Convolutional Networks 2020-08-20	연결 없음
Sudo rm-rf (U=8)	SI-SDR-WB 18.6 PESQ-WB 2.69	Continual self-training with bootstrapped remixing for speech enhancement 2021-10-19	코드 1
Proposed (0.35)	PESQ-NB 2.65 PESQ-WB 2.65	Weighted Speech Distortion Losses for Neural-network-based Real-time Speech Enhancement 2020-02-12	코드 1 · 코드 2 · 코드 3
RemixIT (w Sudo U=32)	SI-SDR-WB 18.0 PESQ-WB 2.60	Continual self-training with bootstrapped remixing for speech enhancement 2021-10-19	코드 1
RemixIT (w Sudo U=32)	SI-SDR-WB 16.0 PESQ-WB 2.34	RemixIT: Continual self-training of speech enhancement models via bootstrapped remixing 2022-02-17	코드 1 · 코드 2
Noisy	SI-SDR-WB 9.1 PESQ-WB 1.58 STOI 91.5	연결 없음 논문 날짜 미기재	연결 없음