Scalable Speech Enhancement with Dynamic Channel Pruning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Miccini, Riccardo, Laroche, Clement, Piechowiak, Tobias, Pezzarossa, Luca
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915268701192192
author Miccini, Riccardo
Laroche, Clement
Piechowiak, Tobias
Pezzarossa, Luca
author_facet Miccini, Riccardo
Laroche, Clement
Piechowiak, Tobias
Pezzarossa, Luca
contents Speech Enhancement (SE) is essential for improving productivity in remote collaborative environments. Although deep learning models are highly effective at SE, their computational demands make them impractical for embedded systems. Furthermore, acoustic conditions can change significantly in terms of difficulty, whereas neural networks are usually static with regard to the amount of computation performed. To this end, we introduce Dynamic Channel Pruning to the audio domain for the first time and apply it to a custom convolutional architecture for SE. Our approach works by identifying unnecessary convolutional channels at runtime and saving computational resources by not computing the activations for these channels and retrieving their filters. When trained to only use 25% of channels, we save 29.6% of MACs while only causing a 0.75% drop in PESQ. Thus, DynCP offers a promising path toward deploying larger and more powerful SE solutions on resource-constrained devices.
format Preprint
id arxiv_https___arxiv_org_abs_2412_17121
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Scalable Speech Enhancement with Dynamic Channel Pruning
Miccini, Riccardo
Laroche, Clement
Piechowiak, Tobias
Pezzarossa, Luca
Audio and Speech Processing
Machine Learning
Sound
Speech Enhancement (SE) is essential for improving productivity in remote collaborative environments. Although deep learning models are highly effective at SE, their computational demands make them impractical for embedded systems. Furthermore, acoustic conditions can change significantly in terms of difficulty, whereas neural networks are usually static with regard to the amount of computation performed. To this end, we introduce Dynamic Channel Pruning to the audio domain for the first time and apply it to a custom convolutional architecture for SE. Our approach works by identifying unnecessary convolutional channels at runtime and saving computational resources by not computing the activations for these channels and retrieving their filters. When trained to only use 25% of channels, we save 29.6% of MACs while only causing a 0.75% drop in PESQ. Thus, DynCP offers a promising path toward deploying larger and more powerful SE solutions on resource-constrained devices.
title Scalable Speech Enhancement with Dynamic Channel Pruning
topic Audio and Speech Processing
Machine Learning
Sound
url https://arxiv.org/abs/2412.17121