Modulating State Space Model with SlowFast Framework for Compute-Efficient Ultra Low-Latency Speech Enhancement

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Cheng, Longbiao, Pandey, Ashutosh, Xu, Buye, Delbruck, Tobi, Ithapu, Vamsi Krishna, Liu, Shih-Chii
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910771894550528
author Cheng, Longbiao
Pandey, Ashutosh
Xu, Buye
Delbruck, Tobi
Ithapu, Vamsi Krishna
Liu, Shih-Chii
author_facet Cheng, Longbiao
Pandey, Ashutosh
Xu, Buye
Delbruck, Tobi
Ithapu, Vamsi Krishna
Liu, Shih-Chii
contents Deep learning-based speech enhancement (SE) methods often face significant computational challenges when needing to meet low-latency requirements because of the increased number of frames to be processed. This paper introduces the SlowFast framework which aims to reduce computation costs specifically when low-latency enhancement is needed. The framework consists of a slow branch that analyzes the acoustic environment at a low frame rate, and a fast branch that performs SE in the time domain at the needed higher frame rate to match the required latency. Specifically, the fast branch employs a state space model where its state transition process is dynamically modulated by the slow branch. Experiments on a SE task with a 2 ms algorithmic latency requirement using the Voice Bank + Demand dataset show that our approach reduces computation cost by 70% compared to a baseline single-branch network with equivalent parameters, without compromising enhancement performance. Furthermore, by leveraging the SlowFast framework, we implemented a network that achieves an algorithmic latency of just 62.5 μs (one sample point at 16 kHz sample rate) with a computation cost of 100 M MACs/s, while scoring a PESQ-NB of 3.12 and SISNR of 16.62.
format Preprint
id arxiv_https___arxiv_org_abs_2411_02019
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Modulating State Space Model with SlowFast Framework for Compute-Efficient Ultra Low-Latency Speech Enhancement
Cheng, Longbiao
Pandey, Ashutosh
Xu, Buye
Delbruck, Tobi
Ithapu, Vamsi Krishna
Liu, Shih-Chii
Audio and Speech Processing
Machine Learning
Sound
Deep learning-based speech enhancement (SE) methods often face significant computational challenges when needing to meet low-latency requirements because of the increased number of frames to be processed. This paper introduces the SlowFast framework which aims to reduce computation costs specifically when low-latency enhancement is needed. The framework consists of a slow branch that analyzes the acoustic environment at a low frame rate, and a fast branch that performs SE in the time domain at the needed higher frame rate to match the required latency. Specifically, the fast branch employs a state space model where its state transition process is dynamically modulated by the slow branch. Experiments on a SE task with a 2 ms algorithmic latency requirement using the Voice Bank + Demand dataset show that our approach reduces computation cost by 70% compared to a baseline single-branch network with equivalent parameters, without compromising enhancement performance. Furthermore, by leveraging the SlowFast framework, we implemented a network that achieves an algorithmic latency of just 62.5 μs (one sample point at 16 kHz sample rate) with a computation cost of 100 M MACs/s, while scoring a PESQ-NB of 3.12 and SISNR of 16.62.
title Modulating State Space Model with SlowFast Framework for Compute-Efficient Ultra Low-Latency Speech Enhancement
topic Audio and Speech Processing
Machine Learning
Sound
url https://arxiv.org/abs/2411.02019