Email this article Improved MFCC Feature Extraction Combining Symmetric ICA Algorithm for Robust Speech Recognition
Abstract
Independent component analysis (ICA), instead of the traditional discrete cosine transform (DCT), is often used to project log Mel spectrum in robust speech feature extraction. The paper proposed using symmetric orthogonalization in ICA for projecting log Mel spectrum into a new feature space as a substitute in extracting speech features to solve the problem of cumulative error and unequal weights that deflation orthogonalization brings, so as to improve the robustness of speech recognition systems, and increase the efficiency of estimation at the same time. Furthermore, the paper studied the nonlinearities of the objective function in ICA and their coefficients, tested them in all kinds of environments, finding that they influenced the recognition rate greatly in speech recognition systems, and applied a new coefficient in the proposed method. Experiments based on HMM and Aurora-2 speech corpus suggested that the new method was superior to deflation-based ICA and MFCC.
Keywords
References
U. Shrawankar and V. Thakare. "Feature Extraction for a Speech Recognition System in Noisy Environment: A Study," Computer Engineering and Applications (ICCEA), 2010 Second International Conference on. 2010.
Z. Tuske, P. Golik, R. Schluter, and F.R. Drepper. "Non-stationary feature extraction for automatic speech recognition," Acoustics, Speech and Signal Processing (ICASSP), 2011 IEEE International Conference on. 2011.
W. Qiang, Z. Liqing, and S. Guangchuan, "Robust Multifactor Speech Feature Extraction Based on Gabor Analysis," Audio, Speech, and Language Processing, IEEE Transactions on, 2011. vol 19(4), pp. 927-936, 2011.
H. Hsieh, J. Chien, K. Shinoda, and S. Furui. "Independent component analysis for noisy speech recognition," Acoustics, Speech and Signal Processing, (ICASSP). IEEE International Conference on. 2009.
E. Ollila, "The Deflation-Based FastICA Estimator: Statistical Analysis Revisited," Signal Processing, IEEE Transactions on, 2010. vol 58(3), pp. 1527-1541, 2011.
A. Hyvarinen, "Fast and robust fixed-point algorithms for independent component analysis," Neural Networks, IEEE Transactions on, 1999. vol 10(3), pp. 626-634, 1999.
X. Zou, P. Jancovic, and M. Kokuer, "On the Effectiveness of the ICA-based signal representation in non-Gaussian Noise," Icsp: 2008 9th International Conference on Signal Processing, vols 1-5, pp. 1-4, 2008.
P. Somervuo, "Experiments with linear and nonlinear feature transformations in HMM based phone recognition," 2003 Ieee International Conference on Acoustics, Speech, and Signal Processing, vol I, pp. 52-55, 2003.
P. Hyunsin, T. Takiguchi, and Y. Ariki. "Integration of Phoneme-Subspaces Using ICA for Speech Feature Extraction and Recognition," Hands-Free Speech Communication and Microphone Arrays, 2008. HSCMA 2008. 2008.
R. Schluter, A. Zolnay, and H. Ney. "Feature combination using linear discriminant analysis and its pitfalls," INTERSPEECH 2006 and 9th International Conference on Spoken Language Processing, INTERSPEECH 2006 - ICSLP, September 17, 2006 - September 21, 2006. 2006. Pittsburgh, PA, United states: DUMMY PUBID. L. Potamitis, N. Fakotakis, and G. Kokkinakis, "Independent component analysis applied to feature extraction for robust automatic speech recognition," Electronics Letters, vol 36(23) pp. 1977-1978, 2000. A. Hyvärinen, J. Karhunen, and E. Oja, Independent component analysis. vol. 26, 2001. J.H. Lee, H.Y. Jung, T.W. Lee, and S.Y. Lee, "Speech feature extraction using independent component analysis," 2000 Ieee International Conference on Acoustics, Speech, and Signal Processing, vols I-Vi, pp.1631-1634, 2000.
P. Tichavský, Z. Koldovský, and E. Oja, "Speed and accuracy enhancement of linear ICA techniques using rational nonlinear functions," Independent Component Analysis and Signal Separation, pp. 285-292, 2007.
Full Text: PDF