Comparison of Multi-Scale Speaker Vectors and S-Vectors for Zero-Shot Speech Synthesis
Abstract
We compare a novel speaker encoder model, called Multi-Scale Speaker (MSS) Vectors, with state-of-the-art s-vectors model for zero-shot speech synthesis. The s-vectors model relies on a modified transformer self-attention network for its architecture. The MSS vectors model introduces a multi-scale approach to the s-vectors model. Results demonstrate that our model produces more natural and similar-sounding synthesized speech for unseen speakers in a zero-shot speech synthesis system.
Department(s)
Computer Science
Document Type
Conference Proceeding
DOI
10.1109/ISM55400.2022.00055
Keywords
speaker adaptation, speaker embedding, speaker encoder, text to speech
Publication Date
1-1-2022
Recommended Citation
Cory, Tristin and Iqbal, Razib, "Comparison of Multi-Scale Speaker Vectors and S-Vectors for Zero-Shot Speech Synthesis" (2022). Faculty Scholarship. 793.
https://bearworks.missouristate.edu/articles00/793
Journal Title
Proceedings 2022 IEEE International Symposium on Multimedia Ism 2022
