Acoustic Modeling Using Deep Belief Networks
- Resource Type
- Periodical
- Authors
- Mohamed, A.; Dahl, G. E.; Hinton, G.
- Source
- IEEE Transactions on Audio, Speech, and Language Processing IEEE Trans. Audio Speech Lang. Process. Audio, Speech, and Language Processing, IEEE Transactions on. 20(1):14-22 Jan, 2012
- Subject
- Signal Processing and Analysis
Communication, Networking and Broadcast Technologies
Hidden Markov models
Data models
Training
Artificial neural networks
Speech
Speech recognition
Computational modeling
Acoustic modeling
deep belief networks (DBNs)
neural networks
phone recognition
- Language
- ISSN
- 1558-7916
1558-7924
Gaussian mixture models are currently the dominant technique for modeling the emission distribution of hidden Markov models for speech recognition. We show that better phone recognition on the TIMIT dataset can be achieved by replacing Gaussian mixture models by deep neural networks that contain many layers of features and a very large number of parameters. These networks are first pre-trained as a multi-layer generative model of a window of spectral feature vectors without making use of any discriminative information. Once the generative pre-training has designed the features, we perform discriminative fine-tuning using backpropagation to adjust the features slightly to make them better at predicting a probability distribution over the states of monophone hidden Markov models.