Discriminative semi-supervised training for keyword search in low resource languages

In this paper, we investigate semi-supervised training for low resource languages where the initial systems may have high error rate (≥ 70.0% word eror rate). To handle the lack of data, we study semi-supervised techniques including data selection, data weighting, discriminative training and multilayer perceptron learning to improve system performance. The entire suite of semi-supervised methods presented in this paper was evaluated under the IARPA Babel program for the keyword spotting tasks. Our semi-supervised system had the best performance in the OpenKWS13 surprise language evaluation for the limited condition. In this paper, we describe our work on the Turkish and Vietnamese systems.

[1]  Jing Huang,et al.  Multi-View and Multi-Objective Semi-Supervised Learning for HMM-Based Automatic Speech Recognition , 2012, IEEE Transactions on Audio, Speech, and Language Processing.

[2]  Richard M. Schwartz,et al.  Score normalization and system combination for improved keyword spotting , 2013, 2013 IEEE Workshop on Automatic Speech Recognition and Understanding.

[3]  Martin Karafi.  BUT BABEL system for spontaneous Cantonese , 2013 .

[4]  Martin Karafiát,et al.  Semi-supervised bootstrapping approach for neural network feature extractor training , 2013, 2013 IEEE Workshop on Automatic Speech Recognition and Understanding.

[5]  Richard M. Schwartz,et al.  Unsupervised versus supervised training of acoustic models , 2008, INTERSPEECH.

[6]  George Saon,et al.  Penalty function maximization for large margin HMM training , 2008, INTERSPEECH.

[7]  Mark J. F. Gales,et al.  Discriminative map for acoustic model adaptation , 2003, 2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03)..

[8]  Mark J. F. Gales,et al.  Unsupervised training and directed manual transcription for LVCSR , 2010, Speech Commun..

[9]  Tanja Schultz,et al.  Generalized Baum-Welch Algorithm and its Implication to a New Extended Baum-Welch Algorithm , 2011, INTERSPEECH.

[10]  Spyridon Matsoukas,et al.  Developing a Speech Activity Detection System for the DARPA RATS Program , 2012, INTERSPEECH.

[11]  Spyridon Matsoukas,et al.  Region Dependent Transform on MLP Features for Speech Recognition , 2011, INTERSPEECH.

[12]  Alexander Gruenstein,et al.  Unsupervised Testing Strategies for ASR , 2011, INTERSPEECH.

[13]  Hermann Ney,et al.  Unsupervised training of acoustic models for large vocabulary continuous speech recognition , 2005, IEEE Transactions on Speech and Audio Processing.

[14]  Mark J. F. Gales,et al.  Unsupervised Training for Mandarin Broadcast News and Conversation Transcription , 2007, 2007 IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07.

[15]  Daniel Povey,et al.  Minimum Phone Error and I-smoothing for improved discriminative training , 2002, 2002 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[16]  Brian Kingsbury,et al.  Boosted MMI for model and feature-space discriminative training , 2008, 2008 IEEE International Conference on Acoustics, Speech and Signal Processing.

[17]  Jan Cernocký,et al.  BUT BABEL system for spontaneous Cantonese , 2013, INTERSPEECH.