Effective Integration of KAN for Keyword Spotting

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Xu, Anfeng, Zhang, Biqiao, Kong, Shuyu, Huang, Yiteng, Yang, Zhaojun, Srivastava, Sangeeta, Sun, Ming
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866913889733574656
author Xu, Anfeng
Zhang, Biqiao
Kong, Shuyu
Huang, Yiteng
Yang, Zhaojun
Srivastava, Sangeeta
Sun, Ming
author_facet Xu, Anfeng
Zhang, Biqiao
Kong, Shuyu
Huang, Yiteng
Yang, Zhaojun
Srivastava, Sangeeta
Sun, Ming
contents Keyword spotting (KWS) is an important speech processing component for smart devices with voice assistance capability. In this paper, we investigate if Kolmogorov-Arnold Networks (KAN) can be used to enhance the performance of KWS. We explore various approaches to integrate KAN for a model architecture based on 1D Convolutional Neural Networks (CNN). We find that KAN is effective at modeling high-level features in lower-dimensional spaces, resulting in improved KWS performance when integrated appropriately. The findings shed light on understanding KAN for speech processing tasks and on other modalities for future researchers.
format Preprint
id arxiv_https___arxiv_org_abs_2409_08605
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Effective Integration of KAN for Keyword Spotting
Xu, Anfeng
Zhang, Biqiao
Kong, Shuyu
Huang, Yiteng
Yang, Zhaojun
Srivastava, Sangeeta
Sun, Ming
Audio and Speech Processing
Sound
Keyword spotting (KWS) is an important speech processing component for smart devices with voice assistance capability. In this paper, we investigate if Kolmogorov-Arnold Networks (KAN) can be used to enhance the performance of KWS. We explore various approaches to integrate KAN for a model architecture based on 1D Convolutional Neural Networks (CNN). We find that KAN is effective at modeling high-level features in lower-dimensional spaces, resulting in improved KWS performance when integrated appropriately. The findings shed light on understanding KAN for speech processing tasks and on other modalities for future researchers.
title Effective Integration of KAN for Keyword Spotting
topic Audio and Speech Processing
Sound
url https://arxiv.org/abs/2409.08605