G-Mamba: Sparse Graph-Guided Mamba for Audio-Visual Speech Enhancement

2026-09-15

AI Summary

本文提出SG-Mamba，一种轻量级音频-视觉语音增强框架，通过稀疏异质图和线性复杂度Mamba骨干解决计算效率与跨模态对齐准确性之间的权衡问题。

Abstract

Lightweight audio-visual speech enhancement (AVSE) models face a critical trade-off between computational efficiency and cross-modal alignment accuracy. While simple concatenation lacks relational expressiveness, dense cross-attention incurs computational overhead and is prone to unreliable cross-modal correspondence under strong acoustic interference. We propose Sparse Graph-Guided Mamba (SG-Mamba), a lightweight AVSE framework that integrates a sparse heterogeneous graph with a linear-complexity Mamba backbone. The graph explicitly models modality-specific relations through content-adaptive attention and cross-frame audio-visual connections, while Mamba captures long-range temporal context. We further introduce an audio skip connection to preserve spectral detail without sacrificing noise suppression. Evaluated on LRS3, SG-Mamba achieves competitive or superior performance against strong lightweight baselines and reaches 13.091 dB SI-SDR under noise-only condition. It also remains robust in cluttered multi-speaker conditions with a competitive cost of 3.45 G MACs (or 6.90 G FLOPs). Results on VoxCeleb2 further suggest that explicit structural priors improve robustness, generalizability, and computational efficiency in lightweight AVSE.

Problem

• lightweight AVSE
• computational efficiency
• cross-modal alignment
• acoustic interference

Innovation

• Sparse Heterogeneous Graph
• Content-Adaptive Attention
• Cross-Frame Audio-Visual Connections
• Audio Skip Connection