Volume 9 Number 10 (Oct. 2014)
Home > Archive > 2014 > Volume 9 Number 10 (Oct. 2014) >
JSW 2014 Vol.9(10): 2706-2712 ISSN: 1796-217X
doi: 10.4304/jsw.9.10.2706-2712

Speeding up deep neural network based speech recognition systems

Yeming Xiao, Yujing Si, Ji Xu, Jielin Pan, Yonghong Yan

Key Laboratory of Speech Acoustics and Content Understanding, Institute of Acoustics, Chinese Academy of Sciences, Beijing, China 100190

Abstract—Recently, deep neural network (DNN) based acoustic modeling has been successfully applied to large vocabulary continuous speech recognition (LVCSR) tasks. A relative word error reduction around 20% can be achieved compared to a state-of-the-art discriminatively trained Gaussian Mixture Model (GMM). However, due to the huge number of parameters in the DNN, real-time decoding is a bottleneck for the DNN based speech recognition systems. In this paper, we adopt several techniques for the speed optimization of the DNN-based system. Specifically, we use singular value decomposition (SVD) to reduce the model parameters, use the SSE instruction sets for the parallel calculation in the data space, and quantize the model parameters reasonably to convert the floating-point arithmetic into fixed-point arithmetic. Besides, taking the characteristics of speech signal into account, we use a frameskipping method when evaluating the posterior probabilities. Finally, compared to the un-optimized baseline system, with negligible recognition performance loss, the decoding realtime factor of the optimized one is significantly reduced, from 6.1 to 0.31. And this response speed can basically meet the requirement of our real applications.

Index Terms—Large Vocabulary Continuous Speech Recognition, Acoustic Modeling, Deep Neural Network, SSE Instructions.

[PDF]

Cite: Yeming Xiao, Yujing Si, Ji Xu, Jielin Pan, Yonghong Yan, "Speeding up deep neural network based speech recognition systems," Journal of Software vol. 9, no. 10, pp. 2706-2712, 2014.

General Information

  • ISSN: 1796-217X (Online)

  • Abbreviated Title: J. Softw.

  • Frequency:  Quarterly

  • APC: 500USD

  • DOI: 10.17706/JSW

  • Editor-in-Chief: Prof. Antanas Verikas

  • Executive Editor: Ms. Cecilia Xie

  • Abstracting/ Indexing: DBLP, EBSCO,
           CNKIGoogle Scholar, ProQuest,
           INSPEC(IET), ULRICH's Periodicals
           Directory, WorldCat, etc

  • E-mail: jsweditorialoffice@gmail.com

  • Jun 12, 2024 News!

    Vol 19, No 2 has been published with online version   [Click]

  • Jan 04, 2024 News!

    JSW will adopt Article-by-Article Work Flow

  • Apr 01, 2024 News!

    Vol 14, No 4- Vol 14, No 12 has been indexed by IET-(Inspec)     [Click]

  • Apr 01, 2024 News!

    Papers published in JSW Vol 18, No 1- Vol 18, No 6 have been indexed by DBLP   [Click]

  • Mar 01, 2024 News!

    Vol 19, No 1 has been published with online version    [Click]