The improved variable step size least mean squares algorithm for speech enhancement

Kecheng Fan; Hao Yang; Qi Wang

doi:10.1117/12.3029307

6 May 2024 The improved variable step size least mean squares algorithm for speech enhancement

Kecheng Fan, Hao Yang, Qi Wang

Proceedings Volume 13107, Fourth International Conference on Sensors and Information Technology (ICSI 2024); 131073V (2024) https://doi.org/10.1117/12.3029307
Event: Fourth International Conference on Sensors and Information Technology (ICSI 2024), 2024, Xiamen, China

Abstract

This paper proposes a new variable step-size adaptive filtering algorithm based on the LMS algorithm and compares its performance with Least Mean Squares(LMS), Normalized Least Mean Squares(NLMS), Modified Sigmoid- LMS(MLMS) and Regularized NLMS(RNLMS) algorithms. The results indicate that compared to the other comparative algorithms, this algorithm can more effectively improve the signal-to-noise ratio of the filter output. Furthermore, within a wide input signal-to-noise ratio range, this algorithm can consistently enhance the filter signal-to-noise ratio output, effectively addressing the problem of inconsistent filtering effects caused by the input signal-to-noise ratio sensitivity in algorithms such as NLMS algorithms.

(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Kecheng Fan, Hao Yang, and Qi Wang "The improved variable step size least mean squares algorithm for speech enhancement", Proc. SPIE 13107, Fourth International Conference on Sensors and Information Technology (ICSI 2024), 131073V (6 May 2024); https://doi.org/10.1117/12.3029307

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available