Research

This page contains some of the research results from my PhD study at Sound and Image Processing Lab, (KTH) Royal Institute of Technology, Stockholm. My field of research was in statistical modelling based approach for speech enhancement and data compression. 

I received my PhD degree in June 2007. 

Matlab code

 
Entropy-Constrained Vector Quantization (ECVQ) using Gaussian Mixture Models (GMM)

Full Matlab source code for the GMM based ECVQ as published in 

- D. Y. Zhao, J. Samuelsson, and M. Nilsson, “On entropy-constrained vector quantization using Gaussian mixture models,” IEEE Trans. Communications, vol. 56, pp. 2094–2104, Dec. 2008.

FreeBSD license. Use on your own risk. 
 Download


Ph.D. thesis

[1]  D. Y. Zhao, "Model Based Speech Enhancement and Coding", Ph.D. thesis, Royal Institute of Technology (KTH), TRITA-EE 2007:018, June 2007.[Download]


Journal publications

[2]  D. Y. Zhao, J. Samuelsson, and M. Nilsson, “On entropy-constrained vector quantization using Gaussian mixture models,” IEEE Trans. Communications, vol. 56, pp. 2094–2104, Dec. 2008. [Download]

[3]  D. Y. Zhao, W. B. Kleijn, A. Ypma, and B. de Vries, “On-line noise estimation using stochastic-gain HMM for speech enhancement,” IEEE Trans. Audio, Speech and Language Processing, vol. 16, pp. 835–846, May 2008. [Download]

[4]  D. Y. Zhao and W. B. Kleijn, “HMM-based gain-modeling for enhancement of speech in noise,” in IEEE Trans. Audio, Speech and Language Processing, vol. 15, pp. 882–892, Mar. 2007. [Download]

[5]  V. Grancharov, D. Y. Zhao, J. Lindblom, and W. B. Kleijn, “Low complexity, non-intrusive speech quality assessment,” in IEEE Trans. Audio, Speech and Language Processing, vol. 14, pp. 1948–1956, Nov. 2006. [Download]


Conference publications

[6]  D.  Y.  Zhao,  J.  Samuelsson,  and  M.  Nilsson,  “GMM-based entropy-constrained vector quantization,” in Proc.   IEEE Int. Conf.  Acoustics, Speech and Signal Processing, Apr.  2007, pp. 1097–1100. [Download]

[7]  V. Grancharov,  D. Y. Zhao,  J. Lindblom,  and W. B. Kleijn, “Low complexity, non-intrusive speech quality assessment,” in IEEE Trans.  Audio, Speech and Language Processing, vol. 14, pp. 1948–1956, Nov. 2006.

[8]  V. Grancharov,  D. Y. Zhao,  J. Lindblom,  and W. B. Kleijn, “Non-intrusive speech quality assessment with low computational complexity,” in Interspeech - ICSLP, September 2006.

[9]  D. Y. Zhao and W. B. Kleijn, “HMM-based speech enhancement using explicit gain modeling,” in Proc.  IEEE Int.  Conf. Acoustics, Speech and Signal Processing, vol. 1, May 2006, pp. 161–164. Student Paper Contest Winner [Download

[10]  D. Y. Zhao and W. B. Kleijn,  “On noise gain estimation for HMM-based speech enhancement,” in Proc.  Interspeech, Sep. 2005, pp. 2113–2116. [Download]

[11]  D. Y. Zhao and W. B. Kleijn,  “Multiple-description vector quantization using translated lattices with local optimization,”  in Proceedings  IEEE  Global Telecommunications Conference, vol. 1, 2004, pp. 41 – 45. [Download]

[12]  J. Plasberg,  D. Y. Zhao,  and W. B. Kleijn,  “The sensitivity matrix for a spectro-temporal auditory model,” in Proc.  12th European Signal Processing Conf. (EUSIPCO), 2004, pp. 1673–1676. Winner of the Young Authors Best Paper Award [Download]


Patents

[1]  Method and Apparatus for Improved Estimation of Non-stationary Noise for Speech Enhancement, filed by GN ReSound, patent no. 06119399.1-224, 08/23/06.