Next Article in Journal
The Triad Hsp60-miRNAs-Extracellular Vesicles in Brain Tumors: Assessing Its Components for Understanding Tumorigenesis and Monitoring Patients
Previous Article in Journal
Comparison of Endurance Time Prediction of Biceps Brachii Using Logarithmic Parameters of a Surface Electromyogram during Low-Moderate Level Isotonic Contractions
Article

Phonetic Variation Modeling and a Language Model Adaptation for Korean English Code-Switching Speech Recognition

1
Software Development Department, IIR TECH Inc., Daejeon 34134, Korea
2
Artificial Intelligence Research Laboratory, Electronics and Telecommunications Research Institute, Daejeon 34129, Korea
*
Author to whom correspondence should be addressed.
Academic Editor: José A. González-López
Appl. Sci. 2021, 11(6), 2866; https://doi.org/10.3390/app11062866
Received: 16 February 2021 / Revised: 16 March 2021 / Accepted: 17 March 2021 / Published: 23 March 2021
(This article belongs to the Section Computing and Artificial Intelligence)
In this paper, we propose a new method for code-switching (CS) automatic speech recognition (ASR) in Korean. First, the phonetic variations in English pronunciation spoken by Korean speakers should be considered. Thus, we tried to find a unified pronunciation model based on phonetic knowledge and deep learning. Second, we extracted the CS sentences semantically similar to the target domain and then applied the language model (LM) adaptation to solve the biased modeling toward Korean due to the imbalanced training data. In this experiment, training data were AI Hub (1033 h) in Korean and Librispeech (960 h) in English. As a result, when compared to the baseline, the proposed method improved the error reduction rate (ERR) by up to 11.6% with phonetic variant modeling and by 17.3% when semantically similar sentences were applied to the LM adaptation. If we considered only English words, the word correction rate improved up to 24.2% compared to that of the baseline. The proposed method seems to be very effective in CS speech recognition. View Full-Text
Keywords: speech recognition; code-switching; language model; domain adaptation; acoustic model; shallow fusion speech recognition; code-switching; language model; domain adaptation; acoustic model; shallow fusion
Show Figures

Figure 1

MDPI and ACS Style

Lee, D.; Kim, D.; Yun, S.; Kim, S. Phonetic Variation Modeling and a Language Model Adaptation for Korean English Code-Switching Speech Recognition. Appl. Sci. 2021, 11, 2866. https://doi.org/10.3390/app11062866

AMA Style

Lee D, Kim D, Yun S, Kim S. Phonetic Variation Modeling and a Language Model Adaptation for Korean English Code-Switching Speech Recognition. Applied Sciences. 2021; 11(6):2866. https://doi.org/10.3390/app11062866

Chicago/Turabian Style

Lee, Damheo, Donghyun Kim, Seung Yun, and Sanghun Kim. 2021. "Phonetic Variation Modeling and a Language Model Adaptation for Korean English Code-Switching Speech Recognition" Applied Sciences 11, no. 6: 2866. https://doi.org/10.3390/app11062866

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop