A deep architecture with bilinear modeling of hidden representations: Applications to phonetic recognition
- 1 March 2012
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- No. 15206149,p. 4805-4808
Abstract
We develop and describe a novel deep architecture, the Tensor Deep Stacking Network (T-DSN), where multiple blocks are stacked one on top of another and where a bilinear mapping from hidden representations to the output in each block is used to incorporate higher-order statistics of the input features. A learning algorithm for the T-DSN is presented, in which the main parameter estimation burden is shifted to a convex sub-problem with a closed-form solution. Using an efficient and scalable parallel implementation, we train a T-DSN to discriminate standard three-state monophones in the TIMIT database. The T-DSN outperforms an alternative pretrained Deep Neural Network (DNN) architecture in frame-level classification (both state and phone) and in the cross-entropy measure. For continuous phonetic recognition, T-DSN performs equivalently to a DNN but without the need for a hard-to-scale, sequential fine-tuning step.Keywords
This publication has 8 references indexed in Scilit:
- Efficient and effective algorithms for training single-hidden-layer neural networksPattern Recognition Letters, 2012
- Scalable stacking and learning for building deep architectures2012
- Boosting attribute and phone estimation accuracies with deep neural networks for detection-based speech recognition2012
- Towards deeper understanding: Deep convex networks for semantic utterance classification2012
- Context-Dependent Pre-Trained Deep Neural Networks for Large-Vocabulary Speech RecognitionIEEE Transactions on Audio, Speech, and Language Processing, 2011
- Acoustic Modeling Using Deep Belief NetworksIEEE Transactions on Audio, Speech, and Language Processing, 2011
- Tensor Decompositions and ApplicationsSIAM Review, 2009
- Stacked generalizationNeural Networks, 1992