Abstract
In this chapter, we describe one of the several possible ways of exploiting deep neural networks (DNNs) in automatic speech recognition systems—the deep neural network-hidden Markov model (DNN-HMM) hybrid system. The DNN-HMM hybrid system takes advantage of DNN’s strong representation learning power and HMM’s sequential modeling ability, and outperforms conventional Gaussian mixture model (GMM)-HMM systems significantly on many large vocabulary continuous speech recognition tasks. We describe the architecture and the training procedure of the DNN-HMM hybrid system and point out the key components of such systems by comparing a range of system setups.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
For the desired segmental model, this duration model is very rough.
- 2.
The independence assumption made in the HMM is one of the reasons why language model weighting is needed. Assuming one doubles the features by extracting a feature for each 5 ms instead of 10 ms, the acoustic model score will be doubled and so the language model weight will also need to be doubled.
- 3.
Unfair comparison was conducted in several papers that compare the hybrid DNN/HMM system and the KL-HMM system. The conclusions in these papers are thus questionable.
References
Aradilla, G., Bourlard, H., Magimai-Doss, M.: Using KL-based acoustic models in a large vocabulary recognition task. In: Proceedings of Annual Conference of International Speech Communication Association (INTERSPEECH), pp. 928–931 (2008)
Aradilla, G., Vepa, J., Bourlard, H.: An acoustic model based on kullback-leibler divergence for posterior features. In: Proceedings of International Conference on Acoustics, Speech and Signal Processing (ICASSP), vol. 4, pp. IV–657 (2007)
Bahl, L., Brown, P., De Souza, P., Mercer, R.: Maximum mutual information estimation of hidden markov model parameters for speech recognition. In: Proceedings of International Conference on Acoustics, Speech and Signal Processing (ICASSP), vol. 11, pp. 49–52 (1986)
Bourlard, H., Morgan, N., Wooters, C., Renals, S.: CDNN: a context dependent neural network for continuous speech recognition. In: Proceedings of International Conference on Acoustics, Speech and Signal Processing (ICASSP), vol. 2, pp. 349–352 (1992)
Bourlard, H., Wellekens, C.J.: Links between Markov models and multilayer perceptrons. IEEE Trans. Pattern Anal. Mach. Intell. (PAMI) 12(12), 1167–1178 (1990)
Dahl, G.E., Yu, D., Deng, L., Acero, A.: Large vocabulary continuous speech recognition with context-dependent DBN-HMMs. In: Proceedings of International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 4688–4691 (2011)
Dahl, G.E., Yu, D., Deng, L., Acero, A.: Context-dependent pre-trained deep neural networks for large-vocabulary speech recognition. IEEE Trans. Audio, Speech Lang. Process. 20(1), 30–42 (2012)
Godfrey, J.J., Holliman, E.: Switchboard-1 Release 2. Linguistic Data Consortium, Philadelphia (1997)
Godfrey, J.J., Holliman, E.C., McDaniel, J.: Switchboard: telephone speech corpus for research and development. In: Proceedings of International Conference on Acoustics, Speech and Signal Processing (ICASSP), vol. 1, pp. 517–520 (1992)
Hennebert, J., Ris, C., Bourlard, H., Renals, S., Morgan, N.: Estimation of global posteriors and forward-backward training of hybrid hmm/ann systems (1997)
Hermansky, H., Ellis, D.P., Sharma, S.: Tandem connectionist feature extraction for conventional HMM systems. In: Proceedings of International Conference on Acoustics, Speech and Signal Processing (ICASSP), vol. 3, pp. 1635–1638 (2000)
Hinton, G., Deng, L., Yu, D., Dahl, G.E., Mohamed, A.R., Jaitly, N., Senior, A., Vanhoucke, V., Nguyen, P., Sainath, T.N., et al.: Deep neural networks for acoustic modeling in speech recognition: the shared views of four research groups. IEEE Signal Process. Mag. 29(6), 82–97 (2012)
Hwang, M., Huang, X.: Shared-distribution hidden Markov models for speech recognition. IEEE Trans. Speech Audio Process. 1(4), 414–420 (1993)
Kapadia, S., Valtchev, V., Young, S.: MMI training for continuous phoneme recognition on the TIMIT database. In: Proceedings of International Conference on Acoustics, Speech and Signal Processing (ICASSP), vol. 2, pp. 491–494 (1993)
Kingsbury, B., Sainath, T.N., Soltau, H.: Scalable minimum bayes risk training of deep neural network acoustic models using distributed hessian-free optimization. In: Proceedings of Annual Conference of International Speech Communication Association (INTERSPEECH) (2012)
Kumar, N., Andreou, A.G.: Heteroscedastic discriminant analysis and reduced rank HMMs for improved speech recognition. Speech Commun. 26(4), 283–297 (1998)
Morgan, N., Bourlard, H.: Continuous speech recognition using multilayer perceptrons with hidden Markov models. In: Proceedings of International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 413–416 (1990)
Morgan, N., Bourlard, H.A.: Neural networks for statistical recognition of continuous speech. Proc. IEEE 83(5), 742–772 (1995)
Ostendorf, M., Digalakis, V.V., Kimball, O.A.: From HMM’s to segment models: a unified view of stochastic modeling for speech recognition. IEEE Trans. Speech Audio Process. 4(5), 360–378 (1996)
Povey, D.: Discriminative Training for Large Vocabulary Speech Recognition. Ph.D. thesis, Cambridge University Engineering Department, Cambridge (2003)
Povey, D., Kanevsky, D., Kingsbury, B., Ramabhadran, B., Saon, G., Visweswariah, K.: Boosted MMI for model and feature-space discriminative training. In: Proceedings of International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 4057–4060 (2008)
Povey, D., Kingsbury, B., Mangu, L., Saon, G., Soltau, H., Zweig, G.: FMPE: discriminatively trained features for speech recognition. In: Proceedings of International Conference on Acoustics, Speech and Signal Processing (ICASSP), vol. 1, pp. 961–964 (2005)
Povey, D., Woodland, P.C.: Minimum phone error and I-smoothing for improved discriminative training. In: Proceedings of International Conference on Acoustics, Speech and Signal Processing (ICASSP), vol. 1, pp. I-105 (2002)
Robinson, A.J., Cook, G., Ellis, D.P., Fosler-Lussier, E., Renals, S., Williams, D.: Connectionist speech recognition of broadcast news. Speech Commun. 37(1), 27–45 (2002)
Seide, F., Li, G., Yu, D.: Conversational speech transcription using context-dependent deep neural networks. In: Proceedings of Annual Conference of International Speech Communication Association (INTERSPEECH), pp. 437–440 (2011)
Senior, A., Heigold, G., Bacchiani, M., Liao, H.: GMM-free DNN training. In: Proceedings of International Conference on Acoustics, Speech and Signal Processing (ICASSP) (2014)
Su, H., Li, G., Yu, D., Seide, F.: Error back propagation for sequence training of context-dependent deep networks for conversational speech transcription. In: Proceedings of International Conference on Acoustics, Speech and Signal Processing (ICASSP) (2013)
Trentin, E., Gori, M.: A survey of hybrid ANN/HMM models for automatic speech recognition. Neurocomputing 37(1), 91–126 (2001)
Yu, D., Deng, L., Dahl, G.: Roles of pre-training and fine-tuning in context-dependent DBN-HMMs for real-world speech recognition. In: Proceedings of Neural Information Processing Systems (NIPS) Workshop on Deep Learning and Unsupervised Feature Learning (2010)
Yu, D., Ju, Y.C., Wang, Y.Y., Zweig, G., Acero, A.: Automated directory assistance system-from theory to practice. In: Proceedings of Annual Conference of International Speech Communication Association (INTERSPEECH), pp. 2709–2712 (2007)
Zhang, B., Matsoukas, S., Schwartz, R.: Discriminatively trained region dependent feature transforms for speech recognition. In: Proceedings of International Conference on Acoustics, Speech and Signal Processing (ICASSP), vol. 1,pp. I–I (2006)
Zhu, Q., Chen, B., Morgan, N., Stolcke, A.: Tandem connectionist feature extraction for conversational speech recognition. In: Machine Learning for Multimodal Interaction, vol. 3361, pp. 223–231. Springer, Berlin (2005)
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2015 Springer-Verlag London
About this chapter
Cite this chapter
Yu, D., Deng, L. (2015). Deep Neural Network-Hidden Markov Model Hybrid Systems. In: Automatic Speech Recognition. Signals and Communication Technology. Springer, London. https://doi.org/10.1007/978-1-4471-5779-3_6
Download citation
DOI: https://doi.org/10.1007/978-1-4471-5779-3_6
Published:
Publisher Name: Springer, London
Print ISBN: 978-1-4471-5778-6
Online ISBN: 978-1-4471-5779-3
eBook Packages: EngineeringEngineering (R0)