Article: Application of support vector machine algorithm in quality evaluation of audio format conversion in music transmission Journal: International Journal of Reasoning-based Intelligent Systems (IJRIS) 2026 Vol.18 No.12 pp.38 - 50 Abstract: In cross-platform music communication, routine audio format conversion is directly affected by coding methods and parameters on listening quality. This paper proposes an SVM-based quality evaluation method for music transmission audio conversion. Select high-quality audio such as WAV and FLAC as reference sources, construct a sample set in common transmission formats such as MP3, AAC, OGG and the bit rate of 96-256 kbps, extract multi-dimensional features such as MFCC, spectral centroid and zero-crossing rate, and form a supervision label by combining subjective listening results. In model construction and verification, prediction performance of different kernel functions is compared and analysed. Experimental results verify that this method precisely captures quality differences under different formats and parameters, with minimal deviation between predicted MOS and subjective scores and stable performance across music types. The results offer technical support for music platform transcoding strategy formulation and transmission quality control. Inderscience Publishers - linking academia, business and industry through research

Title: Application of support vector machine algorithm in quality evaluation of audio format conversion in music transmission

Authors: Haiting Sun

Addresses: College of Music, Weinan Normal University, Xi'an 710061, Shaanxi, China

Abstract: In cross-platform music communication, routine audio format conversion is directly affected by coding methods and parameters on listening quality. This paper proposes an SVM-based quality evaluation method for music transmission audio conversion. Select high-quality audio such as WAV and FLAC as reference sources, construct a sample set in common transmission formats such as MP3, AAC, OGG and the bit rate of 96-256 kbps, extract multi-dimensional features such as MFCC, spectral centroid and zero-crossing rate, and form a supervision label by combining subjective listening results. In model construction and verification, prediction performance of different kernel functions is compared and analysed. Experimental results verify that this method precisely captures quality differences under different formats and parameters, with minimal deviation between predicted MOS and subjective scores and stable performance across music types. The results offer technical support for music platform transcoding strategy formulation and transmission quality control.

Keywords: support vector machine algorithm; music transmission; audio; quality evaluation; MOS.

DOI: 10.1504/IJRIS.2026.153454

International Journal of Reasoning-based Intelligent Systems, 2026 Vol.18 No.12, pp.38 - 50

Received: 26 Dec 2025
Accepted: 03 Feb 2026
Published online: 08 May 2026 *

Title: Application of support vector machine algorithm in quality evaluation of audio format conversion in music transmission

Keep up-to-date