Siamese network contrastive learning model for the objective evaluation of singing sound quality

Read full text Summarise with VIRAT
Authors: Xiaochen Liang
Cited by: 0
References: 24

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

This study extracts target singers' voices from mixed audio with background music and noise, addressing the subjectivity, instability and lack of objective standards in traditional evaluation.An innovative SNN-SpEx+ method is proposed, combining Siamese neural network (SNN) and contrastive learning-based SpEx+.Its parameter-sharing twin architecture unifies the feature space for reference and mixed speech, breaking the bottleneck of feature space dislocation in traditional dual-network structures.Contrastive learning is integrated into vocal extraction to build a 'separation is learning' joint optimisation framework, enhancing adaptability to unknown singers and short reference voices.Experiments on MUSDB18-HQ and NSynth-Singer show SNN-SpEx+ outperforms SpEx++ by 0.85 dB in SI-SDRi and 0.17 in PESQ.For short references (

Fields of Study & Subjects

Music and Audio Processing Speech and Audio Processing Speech Recognition and Synthesis

Publication details

Journal / source
International Journal of Information and Communication Technology
Volume
27
Issue
32
Pages
18–33
Publication date
1 January 2026
Language
EN