Affiliation:
1. Polytechnic University of Timisoara
Abstract
The paper discusses the possibility of phonemes generation based on a recurrent neural network model. In each phoneme a typical or elemental pattern can be identified that repeats itself with slight fluctuations along the signal length. This elemental pattern constitutes the training data for the recurrent neural network. After training, the network can generate three new periods of elemental patterns. In a repetitive loop the network can generate the entire phoneme signal. The model proved very simple and effective, and the generated phonemes gave the impression of a natural sound.
Publisher
Trans Tech Publications, Ltd.
Reference15 articles.
1. J. Holmes and W. Holmes: Speech Synthesis and Recognition, 2nd Edition, Taylor & Francis, N.Y. (2001).
2. D. Jurafsky and J.H. Martin: Speech and Language Processing. An Introduction to Natural Language Processing, Computational Linguistics, and Speech Recognition, Second Edition, Pearson Prentice Hall, (2008).
3. S. McLaughlin and P. Maragos: Nonlinear methods for speech analysis and synthesis. In: Marshall S, Sicuranza G, editor. Advances in nonlinear signal and image processing, Vol. 6. Hindawi Publishing Corporation (2007), p.103.
4. V. Pitsikalis and P. Maragos: Analysis and classification of speech signals by generalized fractal dimension features, Speech Communication, Vol. 51, no. 12, (2009), p.1206–1223.
5. K. Sreenivasa Rao: Role of neural network models for developing speech systems, Sadhana Vol. 36, Part 5, (2011 Oct), p.783–836.
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献
1. Approximation Neural Network for Phoneme Synthesis;Advances in Intelligent Systems and Computing;2015