Methods › Audio › Speech Separation Models › SepFormer
SepFormer
Introduced by Cem Subakan et al. in Attention is All You Need in Speech Separation
archive 2025-07-28 Description, source and code snippet are the archive's method entry.
SepFormer is Transformer-based neural network for speech separation. The SepFormer learns short and long-term dependencies with a multi-scale approach that employs transformers. It is mainly composed of multi-head attention and feed-forward layers. A dual-path framework (introduced by DPRNN) is adopted and RNNs are replaced with a multiscale pipeline composed of transformers that learn both short and long-term dependencies. The dual-path framework enables the mitigation of the quadratic complexity of transformers, as transformers in the dual-path framework process smaller chunks.
The model is based on the learned-domain masking approach and employs an encoder, a decoder, and a masking network, as shown in the figure. The encoder is fully convolutional, while the decoder employs two Transformers embedded inside the dual-path processing block. The decoder finally reconstructs the separated signals in the time domain by using the masks predicted by the masking network.
Papers archive 2025-07-28
12 shown of 12, newest first. Repository counts are the archive's code-links table. A Syntology line states what Syntology ran from that paper's harvested code; it is per sample and not a correctness claim.
-
Beyond Speaker Identity: Text Guided Target Speech Extraction 15 Jan 2025 · 1 repository · arXiv:2501.09169
-
LMAC-TD: Producing Time Domain Explanations for Audio Classifiers 13 Sep 2024 · 0 repositories · arXiv:2409.08655
-
Noise-robust Speech Separation with Fast Generative Correction 11 Jun 2024 · 1 repository · arXiv:2406.07461
-
On Data Sampling Strategies for Training Neural Network Speech Separation Models 14 Apr 2023 · 0 repositories · arXiv:2304.07142
-
Towards Real-Time Single-Channel Speech Separation in Noisy and Reverberant Environments 14 Mar 2023 · 0 repositories · arXiv:2303.07569
-
X-SepFormer: End-to-end Speaker Extraction Network with Explicit Optimization on Speaker Confusion 9 Mar 2023 · 0 repositories · arXiv:2303.05023
-
Separate And Diffuse: Using a Pretrained Diffusion Model for Improving Source Separation 25 Jan 2023 · 0 repositories · arXiv:2301.10752
-
Improving Target Speaker Extraction with Sparse LDA-transformed Speaker Embeddings 16 Jan 2023 · 0 repositories · arXiv:2301.06277
-
Efficient Transformer-based Speech Enhancement Using Long Frames and STFT Magnitudes 23 Jun 2022 · 0 repositories · arXiv:2206.11703
-
Exploring Self-Attention Mechanisms for Speech Separation 6 Feb 2022 · 1 repository · arXiv:2202.02884
-
Monaural source separation: From anechoic to reverberant environments 15 Nov 2021 · 0 repositories · arXiv:2111.07578
-
Attention is All You Need in Speech Separation 25 Oct 2020 · 4 repositories · arXiv:2010.13154
Tasks archive 2025-07-28
11 tasks the archive attaches to papers tagged with this method, by distinct papers. A task without a page in the catalog is plain text.
Usage over time archive 2025-07-28
Components: the archive holds no method-to-method composition, so PwC's Components table cannot be rebuilt; the Papers list carries no Results column for the same reason (the archive does not join its leaderboard rows to method tags).
Categories archive 2025-07-28
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections