Fusion of multiple parameterisations for DNN-based sinusoidal speech synthesis with multi-task learning

Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

Volumn 2015-January, Issue , 2015, Pages 854-858

  Hu, Qiong ^a   Wu, Zhizheng ^a   Richmond, Korin ^a   Yamagishi, Junichi ^a   Stylianou, Yannis ^b   Maia, Ranniery ^b  

^a UNIVERSITY OF EDINBURGH   (United Kingdom)

^b TOSHIBA CORPORATION   (Japan)

Author keywords

Deep neural network; Fusion vocoder; Sinusoidal model; Statistical speech synthesis

Indexed keywords

DECISION TREES; LEARNING SYSTEMS; PARAMETERIZATION; SPEECH; SPEECH SYNTHESIS; VOCODERS;

DEEP NEURAL NETWORKS; HARMONIC AMPLITUDE; MULTITASK LEARNING; SECONDARY TASKS; SINUSOIDAL MODEL; SOURCE FILTERS; STATISTICAL PARAMETRIC SPEECH SYNTHESIS; WAVEFORM GENERATION;

SPEECH COMMUNICATION;

EID: 84959144342     PISSN: 2308457X     EISSN: 19909772     Source Type: Conference Proceeding    
DOI: None     Document Type: Conference Paper

References (29)

1. Y. Agiomyrgiannakis. Vocaine the vocoder and applicationsin speech synthesis. In Proc. ICASSP, 2015.
2. R. Caruna. Multitask learning: A knowledge-based sourceof inductive bias. In Machine Learning: Proceedings ofthe Tenth International Conference, 1993.
3. R. Collobert and J. Weston. A unified architecture fornatural language processing: Deep neural networks withmultitask learning. In Proceedings of the 25th internationalconference on Machine learning, pages 160-167. ACM, 2008.
4. G. Degottex and Y. Stylianou. Analysis and synthesis ofspeech using an adaptive full-band harmonic model. IEEETransactions on Audio, Speech and Language Processing, 21 (10): 2085-2095, 2013.
5. T. Drugman and T. Dutoit. The deterministic plus stochasticmodel of the residual signal and its applications. IEEETransactions on Audio, Speech and Language Processing, 20 (3): 968-981, 2012.
6. D. Erro, I. Sainz, E. Navas, and I. Hernaez. Harmonicsplus noise model based vocoder for statistical parametricspeech synthesis. IEEE Journal of Selected Topics in SignalProcessing, 8 (2): 184-194, 2014.
7. G. Hinton, L. Deng, D. Yu, G. Dahl, A. Mohamed, N. Jaitly, A. Senior, V. Vanhoucke, P. Nguyen, and T. Sainath. Deep neural networks for acoustic modelingin speech recognition: The shared views of four researchgroups. Signal Processing Magazine, IEEE, 29 (6): 82-97, 2012.
8. Q. Hu, K. Richmond, J. Yamagishi, and J. Latorre. Anexperimental comparison of multiple vocoder types. InPro. 8th SSW, 2013.
9. Q. Hu, Y. Stylianou, R. Maia, K. Richmond, and J. Yamagishi. Methods for applying dynamic sinusoidal modelsto statistical parametric speech synthesis. In Proc. ICASSP, 2015.
10. Q. Hu, Y. Stylianou, R. Maia, K. Richmond, J. Yamagishi, and J. Latorre. An investigation of the application of dynamicsinusoidal models to statistical parametric speechsynthesis. In Proc. Interspeech, 2014.
11. Q. Hu, Y. Stylianou, K. Richmond, R. Maia, J. Yamagishi, and J. Latorre. A fixed dimension and perceptually baseddynamic sinusoidal model of speech. In Proc. ICASSP, 2014.
12. Q. Hu, Z. Wu, K. Richmond, J. Yamagishi, Y. Stylianou, R. Maia, S. King, and M. Akamine. Sinusoidal speechsynthesis using deep neural networks. manuscript, 2015.
13. H. Kawahara, I. Masuda-Katsuse, and A. de Cheveigné. Restructuring speech representations using a pitchadaptivetime-frequency smoothing and an instantaneousfrequency-based F0 extraction: Possible role of a repetitivestructure in sounds. Speech communication, 27 (3): 187-207, 1999.
14. J. Kominek, T. Schultz, and A. Black. Synthesizer voicequality of new languages calibrated with mean mel cepstraldistortion. In Pro. SLTU, 2008.
15. Y. Lu, F. Lu, S. Sehgal, S. Gupta, J. Du, C. Tham, P. Green, and V. Wan. Multitask learning in connectionistspeech recognition. In Proceedings of the AustralianInternational Conference on Speech Science and Technology, 2004.
16. S. Parveen and P. Green. Multitask learning in connectionistrobust asr using recurrent neural networks. In INTERSPEECH, 2003.
17. Y. Qian, Y. Fan, W. Hu, and F. Soong. On the training aspectsof deep neural network for parametric tts synthesis. In Proc. ICASSP, 2014.
18. T. Raitio, A. Suni, J. Yamagishi, H. Pulakka, J. Nurminen, M. Vainio, and P. Alku. HMM-based speech synthesisutilizing glottal inverse filtering. IEEE Transactions onAudio, Speech, and Language Processing, 19 (1): 153-165, 2011.
19. K. Richmond, P. Hoole, and S. King. Announcingthe Electromagnetic Articulography (Day 1)Subset of the mngu0 Articulatory Corpus., 2011. http: //dx. doi. org/10. 7488/ds/140.
20. S. Shechtman and A. Sorin. Sinusoidal model parameterizationfor HMM-based TTS system. In Proc. Interspeech, 2010.
21. Y. Stylianou. Harmonic plus noise models for speech, combined with statistical methods, for speech and speakermodification. PhD thesis, Ecole Nationale Supérieure desTélécommunications, 1996.
22. T. Toda and K. Tokuda. A speech parameter generationalgorithm considering global variance for HMM-basedspeech synthesis. IEICE Transactions on Information and Systems, 90 (5): 816-824, 2007.
23. K. Tokuda, T. Yoshimura, T. Masuko, T. Kobayashi, and T. Kitamura. Speech parameter generation algorithms forHMM-based speech synthesis. In Proc. ICASSP, 2000.
24. G. Tur. Multitask learning for spoken language understand ing. In ICASSP, 2006.
25. Z. Wu, C. Botinhao, O. Watts, and S. King. Deep neuralnetworks employing multi-task learning and stacked bottleneckfeatures for speech synthesis. In Proc. ICASSP, 2015.
26. T. Yoshimura, K. Tokuda, T. Masuko, T. Kobayashi, and T. Kitamura. Simultaneous modeling of spectrum, pitchand duration in HMM-based speech synthesis. In Proc. Eurospeech, 1999.
27. H. Zen and A. Senior. Deep mixture density networks foracoustic modeling in statistical parametric speech synthesis. In Proc. ICASSP, 2014.
28. H. Zen, A. Senior, and M. Schuster. Statistical parametricspeech synthesis using deep neural networks. In Proc. ICASSP, 2013.
29. H. Zen, K. Tokuda, and A. Black. Statistical parametricspeech synthesis. Speech Communication, 51 (11): 1039-1064, 2009.