NEURAL NETWORKS FOR VOICE CONVERSION SYSTEM DESIGN
Keywords:
LPC-vocoder, Wavelet Packet Transform, RBF Neural Network, General Regression Neural Network, Robust.Abstract
Voice conversion system formulates conversion function, which can transform specific parameters of source speaker to taregt speaker. In this paper we used voice parameters: shape of the vocal tract, shape of exciatation signal(glottal pulse) and prosodic features(energy, pitch). We compare Radial Basis Function Neural Network and General Regression Neural Network for parameters conversion. We implement new method for outlier detection in dataset.Downloads
References
A.N. Chadha, A comparative performance of various speech analysis-synthesis techniques / A.N. Chadha, J.H. Nirmal, P. Kachare // Int. J. Signal Process. Syst. 2 (1) (2014)
–22.
J. Nirmal, Voice conversion using General Regression Neural Network, Applied Soft Computing / J. Nirmal, M. Zaveri, S. Patnaik, P. Kachare. 2014.
W. Kain Spectral voice conversion for text-to- speech synthesis / W. Kain, M. Macon // In: Proceeding of International Conference on Acoustics, Speech, and Signal Processing, vol. 1, IEEE, 1998, pp. 285–288.
K.S. Rao, Voice conversion by mapping the speakerspecific features using pitch synchronous approach / K.S. Rao // Comput. Speech Lang. 24 (3) (2010). 474–494.
S. Desai Spectral mapping using artificial neural networks for voice conversion/ S. Desai, A.W. Black, B. Yegnanarayana, K. Prahallad // IEEE Trans. Audio Speech Lang.Process. 18 (5) (2010) 954–964.
Sushant V. Rao Novel Pre-processing using Outlier Removal in Voice Conversion/ Sushant V. Rao, Nirmesh J. Shah, Hemant A. Patil, 2016.
S.H. Mohammadi, Voice Conversion Using Deep Neural Networks With Speaker-Independent Pre-training,/ S.H. Mohammadi, A. Kain, 2014.
S.H. Mohammadi Transmutative Voice Conversion / S.H. Mohammadi, A. Kain 2013.
Holmes, J.N. Speech synthesis and recognition/John Holmes and Wendy Holmes.—2 nd ed, 2001.
Набір голосових даних [Електронний ресурс]. – Режим доступу: festvox.org/cmu_arctic.
E. Helander. On the impact ofalignment on voice conversion performance / E. Helander, H Silén, M Gabbouj, 2008.
T.H. Park Introduction To Digital Signal Processing / T.H. Park, 2010.
S. Haykin Neural networks and learning machines / Simon Haykin. – 3rd ed.
M. Bishop Neural Networks for Pattern Recognition / M. Bishop. 1995.
A. Amrouche. Efficient System for Speech Recognition using General Regression Neural Network / A. Amrouche, J.M. Rouvaen, 2008.
Приклади конвертації [Електронний ресурс]. – Режим доступу: https://drive.google.com/open?id=0BwP19oqytjEZVh0Vl9jOEpnaGM.
S. Mondal. Clustering based voiced-unvoicedsilence detection in speech using temporal and spectral parameters/ S. Mondal, A. D. Barman, 2015.
