Analysis of MLP-based hierarchical phoneme posterior probability estimator

IEEE Transactions on Audio, Speech and Language Processing

Volumn 19, Issue 2, 2011, Pages 225-241

  Pinto, Joel ^a   Garimella, Sivaram ^b   Magimai Doss, Mathew ^a   Hermansky, Hynek ^b   Bourlard, Hervé ^a  

^a IDIAP RESEARCH INSTITUTE   (Switzerland)

^b Whiting School of Engineering   (United States)

Author keywords

Hierarchical systems; multilayer perceptrons; posterior probabilities; Volterra series

Indexed keywords

ACOUSTIC FEATURES; CONDITIONAL PROBABILITIES; FEATURE SPACE; HIERARCHICAL ARCHITECTURES; MLP CLASSIFIERS; MULTI LAYER PERCEPTRON; MULTILAYER PERCEPTRONS; PHONEME RECOGNITION; PHONETIC CONFUSIONS; POSTERIOR PROBABILITY; TEMPORAL PATTERN; TRAINING DATA; VOLTERRA SERIES;

CLASSIFIERS; CYBERNETICS; ELECTRIC LOADS; FEATURE EXTRACTION; LINGUISTICS; MULTILAYERS; PATTERN RECOGNITION SYSTEMS; PROBABILITY;

HIERARCHICAL SYSTEMS;

EID: 78049251448     PISSN: 15587916     EISSN: None     Source Type: Journal    
DOI: 10.1109/TASL.2010.2045943     Document Type: Article

References (58)

1. N. Morgan et al., "Pushing the envelope-Aside", IEEE Signal Process. Mag., vol. 22, no. 5, pp. 81-88, Sep. 2005.
2. A. Stolcke et al., "Recent innovations in speech-to-text transcription at SRI-ICSI-UW", IEEE Trans. Audio, Speech, Lang. Process., vol. 14, no. 5, pp. 1729-1744, Sep. 2006.
3. Q. Zhu, B. Chen, N. Morgan, and A. Stolcke, "On using MLP features in LVCSR", in Proc. Interspeech, 2004, pp. 921-924.
4. P. Fousek, L. Lamel, and J.-L. Gauvain, "Transcribing broadcast data using MLP features", in Proc. Interspeech, 2008, pp. 1433-1436.
5. J. Park, F. Diehl, M. Gales, M. Tomalin, and P. Woodland, "Training and adapting MLP features for Arabic speech recognition", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 2009, pp. 4461-4464.
6. H. Bourlard and N. Morgan, Connectionist Speech Recognition-A Hybrid Approach. Norwell, MA: Kluwer, 1994.
7. M. Richard and R. Lippmann, "Neural network classifiers estimate Bayesian a posteriori probabilities", Neural Comput., vol. 3, pp. 461-483, 1991.
8. S. Ikbal, "Nonlinear feature transformations for noise robust speech recognition", Ph. D. dissertation, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, 2004.
9. H. Misra, "Multi-stream processing for noise robust speech recognition", Ph. D. dissertation, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, 2006.
10. H. Hermansky, D. Ellis, and S. Sharma, "Tandem connectionist feature extraction for conventional HMM systems", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 2000, pp. 1635-1638.
11. G. Aradilla, "Acoustic models for posterior features in speech recognition", Ph. D. dissertation, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland, 2008.
12. J. Pinto, B. Yegnanarayana, H. Hermansky, and M. Magimai-Doss, "Exploiting contextual information for improved phoneme recognition", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 2008, pp. 4449-4452.
13. J. Pinto, S. Prasanna, B. Yegnanarayana, and H. Hermansky, "Significance of contextual information in phoneme recognition", Idiap Research Inst., Martigny, Switzerland, Tech. Rep. 28, 2007.
14. S. Boyd, L. O. Chua, and C. A. Desoer, "Analytical foundations of Volterra series", IMA J. Math. Control Inf., vol. 1, pp. 243-282, 1984.
15. V. Volterra, Theory of Functionals and of Integro-Differential Equations. New York: Dover, 1930.
16. D. Ellis and N. Morgan, "Size matters: An empirical study of neural network training for large vocabulary continuous speech recognition", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 1999, vol. 2, pp. 1013-1016.
17. P. Schwarz, P. Matejka, and J. Cernocky, "Hierarchical structures of neural networks for phoneme recognition", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 2006, pp. 325-328.
18. J. Morris and Fosler-Lussier, "Conditional random fields for integrating local discriminative classifiers", IEEE Trans. Audio, Speech, Lang. Process., vol. 16, no. 3, pp. 617-628, 2008.
19. E. Fosler-Lussier and J. Morris, "CRANDEM systems: Conditional random field acoustic models for hidden Markov models", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 2008, pp. 4049-4052.
20. H. Ketabdar and H. Bourlard, "Hierarchical integration of phonetic and lexical knowledge in phone posterior estimation", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 2008, pp. 4065-4068.
21. Y. LeCun, L. Bottou, G. Orr, and K.-R. Muller, G. Orr and K.-R. Muller, Eds., "Efficient BackProp", in Neural Networks: Tricks of the Trade, ser. Lecture Notes in Computer Science. Berlin, Germany: Springer-Verlag, 1998, pp. 9-50, no. 1524.
22. H. Hermansky and S. Sharma, "Temporal patterns (TRAPs) in ASR of noisy speech", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 1999, vol. 1, pp. 289-292.
23. B. Chen, S. Chang, and S. Sivadas, "Learning discriminative temporal patterns in speech: Development of novel TRAPS-like classifiers", in Proc. ICSLP, 2001, pp. 429-432.
24. J. Allen, "How do humans process and recognize speech?", IEEE Trans. Speech. Audio. Process., vol. 2, pp. 567-577, 1994.
25. H. Fletcher, Speech and Hearing in Communication, ASA ed. Melville, NY: Acoustical Society of America, 1995.
26. H. Bourlard and S. Dupont, "A new ASR approach based on independent processing and recombination of partial frequency bands", in Proc. ICSLP, 1996, pp. 422-425.
27. S. Tibrewala and H. Hermansky, "Sub-band based recognition of noisy speech", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 1997, pp. 1255-1258.
28. H. Hermansky and P. Fousek, "Multi-resolution RASTA filtering for tandem based ASR", in Proc. Interspeech, 2005, pp. 361-364.
29. F. Valente and H. Hermansky, "Hierarchical and parallel processing of modulation spectrum for ASR applications", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 2008, pp. 4165-4168.
30. F. Valente and H. Hermansky, "On the combination of auditory and modulation frequency channels for ASR applications", in Proc. Interspeech, 2008, pp. 2242-2245.
31. H. Ketabdar, J. Vepa, S. Bengio, and H. Bourlard, "Using more informative posterior probabilities for speech recognition", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 2006, pp. 29-32.
32. A. Robinson, "An application of recurrent nets to phone probability estimation", IEEE Trans. Neural Netw., vol. 5, no. 2, pp. 298-305, Mar. 1994.
33. S. U. Khan, G. Sharma, and P. Rao, "Speech recognition using neural networks", in Proc. IEEE Conf. Indust. Technol., 2000, vol. 2, pp. 432-437.
34. Y. Abdel-Haleem, "Conditional random fields for continuous speech recognition", Ph. D. dissertation, Univ. of Sheffield, Sheffield, U. K., 2006.
35. F. Grezl, M. Karafiat, S. Kontar, and J. Cernosky, "Probabilistic and bottleneck features for LVCSR of meetings", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 2007, pp. 757-760.
36. D. Povey et al., "FMPE: Discriminatively trained features for speech recognition", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 2005, vol. 1, pp. 961-964.
37. D. Povey and P. Woodland, "Minimum phone error and I-smoothing for improved discriminative training", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 2002, vol. 1, pp. 105-108.
38. K. Sim and M. Gales, "Discriminative semi-parametric trajectory models for speech recognition", Comput. Speech Lang., vol. 21, no. 4, pp. 669-687, 2007.
39. J. Pinto, M. Magimai-Doss, and H. Bourlard, "MLP based hierarchical system for task adaptation in ASR", in Proc. IEEE Workshop Autom. Speech Recognition Understanding (ASRU), 2009, pp. 365-370.
40. K.-F. Lee and H.-W. Hon, "Speaker-independent phone recognition using hidden Markov models", IEEE Trans. Acoust., Speech, Signal Process., vol. 37, no. 11, pp. 1641-1648, Nov. 1989.
41. G. Evermann et al., "Developmentof the 2003 CU-HTK conversational telephone speech transcription system", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 2004, pp. 249-252.
42. P. C. Woodland et al., "The CU-HTK english CTS system", in Proc. Rich Transcription Workshop, 2003 [Online]. Available: http://www.itl.nist.gov/iad/mig/tests/rt/2003-spring/presentations/ctsslides. 2up. letter.pdf
43. J. J. Godfrey, E. C. Holliman, and J. Mcdaniel, "SWITCHBOARD: Telephone speech corpus for research and development", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 1992, vol. 1, pp. 517-520.
44. T. Hain et al., "The development of AMI system for transcription of speech in meetings", in Proc. Mach. Learn. Multimodal Interaction: 2nd Int. Workshop, Revised Sel. Papers, S. Renals and S. Bengio, Eds., 2005, no. 3869, pp. 344-356, Springer-Verlag.
45. S. Fitt, "Documentation and User Guide to UNISYN Lexicon and Post-Lexical Rules", Center for Speech Technol. Res., Univ. of Edinburgh, 2000, Tech. Rep.
46. The ICSI Quicknet Software Package [Online]. Available: http://www.icsi.berkeley.edu/Speech/qn.html
47. SRILM-The SRI Language Modeling Toolkit [Online]. Available: http://www.speech.sri.com/projects/srilm
48. D. Moore et al., "Juicer: A weighted finite state transducer speech decoder", in Proc. Mach. Learn. for Multimodal Interaction: 3rd Int. Workshop, Revised Sel. Papers, S. Renals, S. Bengio, and J. Fiscus, Eds., 2006, no. 4299, pp. 285-296, Springer-Verlag.
49. J. Pinto, G. Sivaram, H. Hermansky, and M. Magimai-Doss, "Volterra series for analyzing MLP based phoneme posterior estimator", Idiap Res. Inst., Tech. Rep. 69, 2008.
50. J. Pinto, G. Sivaram, H. Hermansky, and M. Magimai-Doss, "Volterra series for analyzing MLP based phoneme posterior estimator", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 2009, pp. 1813-1816.
51. V. Marmarelis, Nonlinear Dynamic Modeling of Physiological Systems. New York: Wiley, 2004.
52. W. Gray and B. Nabet, "Volterra series analysis and synthesis of a neural network for velocity estimation", IEEE Trans. Syst., Man, Cybern., vol. 29, no. 2, pp. 190-197, Apr. 1999.
53. W. Knecht, "Nonlinear noise filtering and beamforming using the perceptron and its Volterra approximation", IEEE Trans. Speech Audio Process., vol. 2, no. 1, pp. 55-62, Jan. 1994.
54. G. Stegmayer, "Volterra series and neural networks to model an electronic device nonlinear behavior", in Proc. IEEE Conf. Neural Netw., 2004, vol. 4, pp. 2907-2910.
55. D. Ellis, R. Singh, and Sivadas, "Tandem acoustic modeling in largevocabulary recognition", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 2001, vol. 1, pp. 517-520.
56. S. Sivadas and H. Hermansky, "Hierarchical tandem feature extraction", in Proc. IEEE Conf. Acoust., Speech, Signal Process. (ICASSP), 2002, vol. 1, pp. 809-812.
57. J. Pinto and H. Hermansky, "Combining evidence from a generative and a discriminative model in phoneme recognition", in Proc. Interspeech, 2008, pp. 2414-2417.
58. M. Lehtonen, P. Fousek, and H. Hermansky, "Hierarchical approach for spotting keywords", Idiap Research Inst., Tech. Rep. 41, 2005.