Yi-Jian Wu

Research Interests

Yi-Jian Wu is interested in speech and acoustic processing, in particular speech synthesis. His research topics include phonetic segmentation, speech synthesis and voice conversion. His research aims to realize the speech synthesis with high quality and flexibility.

Education

• Sep. 1996 - Jul. 2001
  • University of Science and Technology of China (USTC)
    • B.E. degree in Electronic Engineering and Infomation Science
• Sep. 2001 - Jul. 2006
  • University of Science and Technology of China (USTC)
    • M.E. degree in Electronic Engineering and Infomation Science, 2003
    • Ph.D. degree in Electronic Engineering and Infomation Science, 2006

Professional Experience

• Apr. 2003 - Mar. 2004
  • Spoken Language Translation Research Laboratories, Advanced Telecommunications Research Institute International, (ATR-SLT), Japan
    • Intern Student
• Jul. 2006 - May. 2007
  • Speech Group, Microsoft Research Asia (MSRA), China
    • Associate Researcher
• May. 2007 - Present
  • Nagoya Institue of Technology, Japan
    • Postdoctoral Research Associate

Publications

Journal Papers

1. W. Guo, Y.-J. Wu and R.H. Wang, "A smoothing method for voiced units concatenation based on time-domain unit fusion," (in Chinese), Jurnal of Chinese Information Processing, Vol. 20, No. 5, pp. 71-76, Sep. 2006.

2. Y.-J. Wu and R.H. Wang, "HMM-based trainable speech synthesis for Chinese," (in Chinese), Jurnal of Chinese Information Processing, Vol. 20, No. 4, pp. 75-81, Jul. 2006.

3. Y.-J. Wu, H. Kawai, J. Ni and R.H. Wang, "Discriminative training and explicit duration modeling for HMM-based automatic segmentation," Speech Communication, Vol. 47, No 4, pp. 397-410, Dec. 2005.

International Conferences

1. Y.-J. Wu and K. Tokuda, "Minimum generation error training by using original spectrum as reference for log spectral distortion measure," Proc. of ICASSP 2009 (accepted), Taiwan, Apr. 2009.

2. H. Lu, Y.-J. Wu, K. Tokuda, L.-R. Dai and R.-H. Wang, "Full covariance state duration modeling for HMM-based speech synthesis," Proc. of ICASSP 2009 (accepted), Taiwan, Apr. 2009.

3. Y.-J. Wu, S. King and K. Tokuda, "Cross-lingual speaker adaptation for HMM-based speech synthesis," Proc. of ISCSLP 2008, Kunming, China, Dec. 2008.

4. L. Qin, Y.-J. Wu, Z.-H. Ling and R.-H. Wang, "Model adaptation for HMM-based speech synthesis under minimum generation error criterion," Proc. of IEEE International Symposium on Multimedia (ISM 2008), Berkeley, California, Dec. 2008.

5. Z.-P. Yu, Y.-J. Wu, H. Zen, Y. Nankaku and K. Tokuda, "Analysis of stream-dependent tying structure for HMM-based speech synthesis," Proc. of ICSP 2008, Beijing, China, Oct. 2008.

6. Y.-J. Wu and K. Tokuda, "Minimum generation error training with direct log spectral distortion on LSPs for HMM-based speech synthesis," Proc. of Interspeech 2008, pp. 577-580, Brisbane, Australia, Sep. 2008.

7. J. Yamagishi, H. Zen, Y.-J. Wu, T. Toda and K. Tokuda, "The HTS-2008 system: yet another evaluation of the speaker-adaptive HMM-based speech synthesis system in the 2008 Blizzard Challenge," Blizzard Challenge 2008 Workshop, Brisbane, Australia, Sep. 2008.

8. L. Qin, Y.-J. Wu, Z.-H. Ling, R.-H. Wang, and L.-R. Dai, "Minimum generation error lineal regression based model adaptation for HMM-based speech synthesis," Proc. of ICASSP 2008, pp. 3953-3956, Las Vegas, USA, Mar. 2008.

9. Y.-J. Wu, H. Zen, Y. Nankaku and K. Tokuda, "Minimum generation error criterion considering global/local variance for HMM-based speech synthesis," Proc. of ICASSP 2008, pp. 4621-4624, Las Vegas, USA, Mar. 2008.

10. L. Ma, Y.-J. Wu, P. Liu and F. Soong, "A MSD-HMM approach to pen trajectory modeling for online handwriting recognition," Proc. of ICDAR 2007, vol. 1, pp. 128-132, Parana, Brazil, Sep. 2007.

11. Y.-J. Wu, R.H. Wang and F. Soong, "Full HMM training for minimizing generation error in synthesis," Proc. of ICASSP 2007, pp. 517-520, Hawaii, USA, Apr. 2007.

12. L. Qin, Z.H. Ling, Y.-J. Wu, B.F. Zhang and R.H. Wang, "HMM-Based Emotional Speech Synthesis Using Average Emotion Model," Proc. of ISCSLP 2006, pp. 233-240, Singapore, Dec. 2006.

13. Z.H. Ling, Y.-J. Wu, Y.P. Wang, L. Qin and R.H. Wang, "USTC system for Blizzard Challenge 2006 an improved HMM-based speech synthesis method, " Blizzard Challenge 2006, Pittsburgh, USA, Sep. 2006

14. L. Qin, Y.-J. Wu, Z.H. Ling and R.H. Wang , "Improving the performance of HMM-Based voice conversion using context clustering decision tree and appropriate regression matrix format", Proc. of Interspeech 2006, pp. 2250-2253, Pittsburgh, USA, Sep. 2006.

15. Y.-J. Wu, W. Guo and R.H. Wang, "Minimum generation error criterion for tree-based clustering of context dependent HMMs", Proc. of Interspeech 2006, pp. 2046-2049, Pittsburgh, USA, Sep. 2006.

16. Y.-J. Wu and R.H. Wang, "Minimum generation error training for HMM-based speech synthesis", in ICASSP 2006, vol. 1, pp. 89-92, Toulouse, France, May. 2006.

17. Y.-J. Wu, H. Kawai, J. Ni and R.H. Wang, "A study on automatic detection of Japanese vowel devoicing for speech synthesis", Proc. of ICSLP2004, pp. 2721-2724, Jeju, Korea, Oct. 2004.

18. Y.-J. Wu, H. Kawai, J. Ni and R.H. Wang, "Minimum segmentation error based discriminative training for speech synthesis application", Proc. of ICASSP2004, pp. 629-632, Mentreal, Quebec, Canada, May 2004.

19. G.P. Chen, Y. Hu, Y.-J. Wu and R.H. Wang, "A concatenative-tone model with its parameters' extraction", Proc. of SP2004, pp. 455-458, Nara, Japan, Mar. 2004

20. L. Sun, Y. Hu, R.H. Wang, and Y.-J. Wu, "A study on duration compensation in Mandarin Chinese", Proc. of SP2004, pp. 239-242, Nara, Japan, Mar. 2004

21. Y.-J. Wu, Y. Hu, X. Wu and R.H. Wang, "A new method of building decision tree based on target information", Proc. of ICSLP2002, pp. 129-132, Denver, Colorado, USA, Sep. 2002.

Domestic Conferences

1. Y.-J. Wu and K. Tokuda, "HMM training by minimizing log spectral distortion between generated and original LSPs for speech synthesis," Proc. of Autumn Meeting of ASJ, 1-4-6, Sep. 2008

2. Y.-J. Wu, H. Zen, Y. Nankaku, and K. Tokuda, "Evaluation of parameter optimization methods for minimum generation error based HMM training," Proc. of Autumn Meeting of ASJ, 3-4-10, Sep. 2007

3. L. Qin, Y.-J. Wu, and R.H. Wang, "Regression matrix tying method for HMM-based voice conversion," the 8th meeting on human-machine speech communication, pp. 298-301, China, Oct. 2005.

4. Y.-J. Wu, H. Kawai, J. Ni and R.H. Wang, "Automatic detection of Japanese vowel devoicing," ASJ Spring Meeting, pp. 387-388, Japan, Apr. 2004.

5. Y.-J. Wu, H. Kawai, J. Ni and R.H. Wang, "HMM-based phonetic segmentation based on discriminative training and explicit duration modeling," ASJ Autumn Meeting, pp. 269-270, Nagoya, Japan, Sep. 2003.

6. Y.-J. Wu, H. Kawai, J. Ni and R.H. Wang, "Minimum segmentation error based discriminative training of HMM for automatic segmenation," SP2003, Technique report of IEICE, pp. 13-18, Hokaido, Japan, Aug. 2003.

Others

1. Y.-J. Wu, "Research on HMM-based speech synthesis," Ph.D. thesis, University of Science and Technology of China, Jun. 2006.

2. Y.-J. Wu, H. Kawai, J. Ni and R.H. Wang, "HMM-based application for speech synthesis," ATR Technical report, Japan, Mar. 2004.