Omni-directional attention mechanism based on Mamba for speech separation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Xue, Ke, Sun, Chang, Fan, Rongfei, Wang, Jing, Hu, Han
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912843336515584
author Xue, Ke
Sun, Chang
Fan, Rongfei
Wang, Jing
Hu, Han
author_facet Xue, Ke
Sun, Chang
Fan, Rongfei
Wang, Jing
Hu, Han
contents Mamba, a selective state-space model (SSM), has emerged as an efficient alternative to Transformers for speech modeling, enabling long-sequence processing with linear complexity. While effective in speech separation, existing approaches, whether in the time or time-frequency domain, typically decompose the input along a single dimension into short one-dimensional sequences before processing them with Mamba, which restricts it to local 1D modeling and limits its ability to capture global dependencies across the 2D spectrogram. In this work, we propose an efficient omni-directional attention (OA) mechanism built upon unidirectional Mamba, which models global dependencies from ten different directions on the spectrogram. We expand the proposed mechanism into two baseline separation models and evaluate on three public datasets. Experimental results show that our approach consistently achieves significant performance gains over the baselines while preserving linear complexity, outperforming existing state-of-the-art (SOTA) systems.
format Preprint
id arxiv_https___arxiv_org_abs_2601_16603
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Omni-directional attention mechanism based on Mamba for speech separation
Xue, Ke
Sun, Chang
Fan, Rongfei
Wang, Jing
Hu, Han
Sound
Audio and Speech Processing
Mamba, a selective state-space model (SSM), has emerged as an efficient alternative to Transformers for speech modeling, enabling long-sequence processing with linear complexity. While effective in speech separation, existing approaches, whether in the time or time-frequency domain, typically decompose the input along a single dimension into short one-dimensional sequences before processing them with Mamba, which restricts it to local 1D modeling and limits its ability to capture global dependencies across the 2D spectrogram. In this work, we propose an efficient omni-directional attention (OA) mechanism built upon unidirectional Mamba, which models global dependencies from ten different directions on the spectrogram. We expand the proposed mechanism into two baseline separation models and evaluate on three public datasets. Experimental results show that our approach consistently achieves significant performance gains over the baselines while preserving linear complexity, outperforming existing state-of-the-art (SOTA) systems.
title Omni-directional attention mechanism based on Mamba for speech separation
topic Sound
Audio and Speech Processing
url https://arxiv.org/abs/2601.16603