Abstract
This paper proposes a learning speed improvement using multi-GPUs on DNN-based acoustic model training in Korean intelligent personal assistant (IPA). DNN learning involves iterative, stochastic parameter updates. These updates depend on the previous updates. The proposed method provides a distributed computing for DNN learning. DNN-based acoustic models are trained by using 320 h length Korean speech corpus. It was shown that the learning speed becomes five times faster on this implementation while maintaining speech recognition rate.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Allcock B, Bester J, Bresnahan J, Chervenak AL, Foster I, Kesselman C, Meder S, Nefedova V, Quesnel D, Tuecke S (2002) Data management and transfer in high performance computational grid environments. Parallel Comput J 28:749–771
Bourlard H, Morgan N (1994) Connectionist speech recognition: a hybrid approach. Kluwer Academic Publishers, pp 1–263. doi:10.1007/978-1-4615-3210-1
Castellano M, Mastronardi G, Tarricone G (2009) Intrusion detection using neural networks: a grid computing based data mining approach. Neural Inf Process 5864:777–785
Deng L, Acero A, Wang Y, Wang K, Hon H, Droppo J, Mahajan M, Huang XD (2002) A speech-centric perspectives for human–computer interface. In: IEEE Work. Multimed. Signal Process, pp 263–267
Dragan RV (2002) Sun one grid engine enterprise edition software, 1 October, pp 1–3
Forster I, Kesselan C (1999) The grid: blueprint for a new computing infrastructure. Morgan Kaufman, San Francisco
Li X, Wang YY, Tur G (2011) Multi-task learning for spoken language understanding with shared slots. In: Proc. Annu. Conf. Int. Speech Commun. Assoc. (INTERSPEECH 2011), pp 701–704
Oh K-S, Jung K (2004) GPU implementation of neural networks. Pattern Recognit 37:1311–1314. doi:10.1016/j.patcog.2004.01.013
Raina R, Madhavan A, Ng AY (2009) Large-scale deep unsupervised learning using graphics processors. In: Proc. 26th Int. Conf. Mach. Learn, pp 873–880
Reed DA, Mendes CL, Lu C, Foster I, Kesselmann C (2003) The grid 2: blueprint for a new computing infrastructure. Morgan Kaufman, San Francisco
Seltzer ML, Droppo J (2013) Multi-task learning in deep neural networks for improved phoneme recognition. In: Proc. 2013 IEEE Int. Conf. Acoust. Speech Signal Process. (ICASSP), pp 6965–6969
Steinkraus D, Buck I, Simard PY (2005) Using GPUs for machine learning algorithms. In: Proc. Eighth Int. Conf. Doc. Anal. Recognit, pp 1115–1120
Torresen O, Landsverk J (1998) A review of parallel implementations of backpropagation neural networks. In: Parallel Archit. Artif. Neural Networks, pp 25–64
Tur G (2006) Multitask learning for spoken language understanding. In: Proc. 2006 IEEE Int. Conf. Acoust. Speech Signal Process. (ICASSP), pp 585–588
Wang G (2014) Context-dependent acoustic modelling for speech recognition. PhD Thesis, National University of Singapore
Acknowledgements
This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and Future Planning (No. NRF-2014R1A1A1002197).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2015 Springer International Publishing Switzerland
About this chapter
Cite this chapter
Lee, D., Kim, KH., Kang, HE., Wang, SH., Park, SY., Kim, J.H. (2015). Learning Speed Improvement Using Multi-GPUs on DNN-Based Acoustic Model Training in Korean Intelligent Personal Assistant. In: Lee, G., Kim, H., Jeong, M., Kim, JH. (eds) Natural Language Dialog Systems and Intelligent Assistants. Springer, Cham. https://doi.org/10.1007/978-3-319-19291-8_27
Download citation
DOI: https://doi.org/10.1007/978-3-319-19291-8_27
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-19290-1
Online ISBN: 978-3-319-19291-8
eBook Packages: Computer ScienceComputer Science (R0)