This paper reports our work on building up a Cantonese Speech-to-Text (STT) system with a syllable based acoustic model. This is a part of an effort in building a STT system to aid dyslexic students who have cognitive deficiency in writing skills but have no problem expressing their ideas through speech. For Cantonese speech recognition, the basic unit of acoustic models can either be the conventional Initial-Final (IF) syllables, or the Onset-Nucleus-Coda (ONC) syllables where finals are further split into nucleus and coda to reflect the intra-syllable variations in Cantonese. By using the Kaldi toolkit, our system is trained using the stochastic gradient descent optimization model with the aid of GPUs for the hybrid Deep Neural Network and Hidden Markov Model (DNN-HMM) with and without I-vector based speaker adaptive training technique. The input features of the same Gaussian Mixture Model with speaker adaptive training (GMM-SAT) to DNN are used in all cases. Experiments show that the ONC-based syllable acoustic modeling with I-vector based DNN-HMM achieves the best performance with the word error rate (WER) of 9.66% and the real time factor (RTF) of 1.38812.
|Title of host publication||Proceedings of the 10th International Conference on Language Resources and Evaluation, LREC 2016|
|Editors||Nicoletta CALZOLARI, Khalid CHOUKRI, Helene MAZO, Asuncion MORENO, Thierry DECLERCK, Sara GOGGI, Marko GROBELINK, Jan ODIJK, Stelios PIPERIDIS, Bente MAEGAARD, Joseph MARIANI|
|Publisher||European Language Resources Association (ELRA)|
|Number of pages||7|
|Publication status||Published - 2016|
|Event||10th International Conference on Language Resources and Evaluation, LREC 2016 - Portoroz, Slovenia|
Duration: 23 May 2016 → 28 May 2016
|Name||Proceedings of the 10th International Conference on Language Resources and Evaluation, LREC 2016|
|Conference||10th International Conference on Language Resources and Evaluation, LREC 2016|
|Period||23/05/16 → 28/05/16|
Bibliographical noteFunding Information:
The project is partially supported by the Hong Kong Government’s ITF fund (ITS/072/14) and the Hong Kong Polytechnic University project RTVU. Our special thanks to all Kaldi community.
- Cantonese speech recognition
- Kaldi toolkit
- Onset-nucleus-coda syllable scheme