Joint Modeling and KAFusion Feature Fusion for Prosody-Controllable Speech Synthesis
Electronics · 25.03.2026 · 15 (7) · 1354
DOI: 10.3390/electronics15071354 ↗Kurz gefasst
Fachartikel in der Zeitschrift Electronics, online erschienen am 25.03.2026. Verfasst von Dongfeng Ye und 3 weiteren Autor:innen. Die Arbeit nennt 33 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Major Project of Xiangjiang Laboratory, Major Scientific Research Project of Hunan Provincial Department of Education, Hunan Provincial Innovation Foundation for Postgraduate.
Abstract (Originalsprache)
To address the limited expressiveness in current speech synthesis caused by coarse-grained prosody modeling and simplistic feature fusion strategies, a joint prosody modeling framework and a nonlinear fusion method named KAFusion are proposed, based on the Kolmogorov–Arnold (KA) representation theorem. The joint modeling integrates pitch and energy as prosodic priors with text encodings to jointly guide duration prediction, enabling explicit control over speech rate and tone. During feature fusion, KAFusion facilitates nonlinear interactions among features through its nested inner and outer functions. Information entropy serves as the quantitative metric, and both theoretical and experimental results demonstrate the fusion module’s efficacy in suppressing redundancy while preserving task-critical content. Evaluations on the AISHELL3 dataset show a 5.8% improvement in MOS over the baseline. Ablation studies further validate the effectiveness of the proposed components, where KAFusion achieves an output entropy of 3.47, which is 18.4% higher than that of linear fusion (2.93) and indicates richer information content.
Förderung
- Major Project of Xiangjiang Laboratory
- Major Scientific Research Project of Hunan Provincial Department of Education
- Hunan Provincial Innovation Foundation for Postgraduate
Zitieren
APA
Ye, D., Jiang, L., Ni, N., & Wan, W. (2026). Joint Modeling and KAFusion Feature Fusion for Prosody-Controllable Speech Synthesis. Electronics, 15(7), 1354. https://doi.org/10.3390/electronics15071354
Vancouver
Ye D, Jiang L, Ni N, Wan W. Joint Modeling and KAFusion Feature Fusion for Prosody-Controllable Speech Synthesis. Electronics. 2026;15(7):1354. doi:10.3390/electronics15071354
BibTeX
@article{Ye2026,
author = {Ye, Dongfeng and Jiang, Lin and Ni, Nianxin and Wan, Wei},
title = {Joint Modeling and KAFusion Feature Fusion for Prosody-Controllable Speech Synthesis},
journal = {Electronics},
year = {2026},
volume = {15},
number = {7},
pages = {1354},
doi = {10.3390/electronics15071354}
}
EUVIMED