The THUEE system for the openKWS14 keyword search evaluation

Author

Meng Cai ; Zhiqiang Lv ; Beili Song ; Yongzhe Shi ; Weilan Wu ; Cheng Lu ; Wei-Qiang Zhang ; Jia Liu

Author_Institution

Dept. of Electron. Eng., Tsinghua Univ., Beijing, China

fYear

2015

fDate

19-24 April 2015

Firstpage

4734

Lastpage

4738

Abstract

The OpenKWS14 keyword search evaluation is one of the most challenging and influential evaluations in the field of speech recognition. Its goal is to build a high-performance keyword search system for a minority language with limited training data in a short period of time. We present the system of the Department of Electronic Engineering, Tsinghua University (THUEE team) for the OpenKWS14 keyword search evaluation. The highlights of the system include the use of convolutional maxout neural networks for acoustic modeling and the use of neural network language models for one-pass lattice generation. The final system is a fusion of 8 sub-systems. The system has achieved an actual term weighted value (ATWV) of 0.5107 for the full language pack (FullLP) condition in the evaluation, ranking third among the participating teams.

Keywords

convolution; neural nets; speech recognition; ATWV; Department of Electronic Engineering; FullLP condition; OpenKWS14 keyword search evaluation; Tsinghua University; acoustic modeling; actual term weighted value; convolutional maxout neural network; full language pack; limited training data; neural network language model; one-pass lattice generation; speech recognition; Acoustics; Artificial neural networks; Biological neural networks; Keyword search; Speech; Speech recognition; Training; Acoustic Modeling; Deep Neural Network; Keyword Spotting; Language Modeling; Low-Resource;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech and Signal Processing (ICASSP), 2015 IEEE International Conference on

Conference_Location

South Brisbane, QLD

Type

conf

DOI

10.1109/ICASSP.2015.7178869

Filename

7178869