Research Article
Speech Signal Feature Extraction Method of Tibetan Speech Synthesis System Based on Machine Learning
4 downloads
@INPROCEEDINGS{10.1007/978-3-030-94551-0_37, author={Ze-guo Liu}, title={Speech Signal Feature Extraction Method of Tibetan Speech Synthesis System Based on Machine Learning}, proceedings={Advanced Hybrid Information Processing. 5th EAI International Conference, ADHIP 2021, Virtual Event, October 22-24, 2021, Proceedings, Part I}, proceedings_a={ADHIP}, year={2022}, month={1}, keywords={Machine learning Tibetan Synthesis system Signal extraction}, doi={10.1007/978-3-030-94551-0_37} }- Ze-guo Liu
Year: 2022
Speech Signal Feature Extraction Method of Tibetan Speech Synthesis System Based on Machine Learning
ADHIP
Springer
DOI: 10.1007/978-3-030-94551-0_37
Abstract
In order to improve the accuracy of Tibetan speech synthesis, a feature extraction method of Tibetan speech synthesis system based on machine learning is proposed. Based on the analysis of Tibetan speech text content, the construction of speech synthesis system is realized. By judging the level of Tibetan prosody, a synthetic encoder is designed to realize the feature extraction of Tibetan speech signal. According to the experimental results, under the condition of normal speaking speed and identical Tibetan speech content, the Tibetan speech synthesized by the speech signal feature extraction method of Tibetan speech synthesis system based on machine learning is more accurate.
Copyright © 2021–2025 ICST
