Faculty of Information Technology, BUT

Publication Details

Lightly supervised vs. semi-supervised training of acoustic model on Luxembourgish for low-resource automatic speech recognition

VESELÝ Karel, PERALES Carlos Segura, SZŐKE Igor, LUQUE Jordi and ČERNOCKÝ Jan. Lightly supervised vs. semi-supervised training of acoustic model on Luxembourgish for low-resource automatic speech recognition. In: Proceedings of Interspeech 2018. Hyderabad: International Speech Communication Association, 2018, pp. 2883-2887. ISSN 1990-9772. Available from: https://www.isca-speech.org/archive/Interspeech_2018/abstracts/2361.html
Czech title
Trénování akustického modelu lucemburštiny pro automatické rozpoznávání řeči s omezenými zdroji s lehkou supervizí vs. bez supervize
Type
conference paper
Language
english
Authors
Veselý Karel, Ing., Ph.D. (DCGM FIT BUT)
Perales Carlos Segura (Telefónica)
Szőke Igor, Ing., Ph.D. (DCGM FIT BUT)
Luque Jordi (Telefónica)
Černocký Jan, doc. Dr. Ing. (DCGM FIT BUT)
URL
Keywords
Luxembourgish, call centers, speech recognition, low-resourced ASR, unsupervised training
Abstract
In this work, we focus on exploiting inexpensive data in order to to improve the DNN acoustic model for ASR. We explore two strategies: The first one uses untranscribed data from the target domain. The second one is related to the proper selection of excerpts from imperfectly transcribed out-of-domain public data, as parliamentary speeches. We found out that both approaches lead to similar results, making them equally beneficial for practical use. The Luxembourgish ASR seed system had a 38.8% WER and it improved by roughly 4% absolute, leading to 34.6% for untranscribed and 34.9% for lightlysupervised data. Adding both databases simultaneously led to 34.4% WER, which is only a small improvement. As a secondary research topic, we experiment with semi-supervised state-level minimum Bayes risk (sMBR) training. Nonetheless, for sMBR we saw no improvement from adding the automatically transcribed target data, despite that similar techniques yield good results in the case of cross-entropy (CE) training.
Published
2018
Pages
2883-2887
Journal
Proceedings of Interspeech, vol. 2018, no. 9, ISSN 1990-9772
Proceedings
Proceedings of Interspeech 2018
Conference
Interspeech 2018, Hyderabad, India, IN
Publisher
International Speech Communication Association
Place
Hyderabad, IN
DOI
BibTeX
@INPROCEEDINGS{FITPUB11844,
   author = "Karel Vesel\'{y} and Segura Carlos Perales and Igor Sz\H{o}ke and Jordi Luque and Jan \v{C}ernock\'{y}",
   title = "Lightly supervised vs. semi-supervised training of acoustic model on Luxembourgish for low-resource automatic speech recognition",
   pages = "2883--2887",
   booktitle = "Proceedings of Interspeech 2018",
   journal = "Proceedings of Interspeech",
   volume = 2018,
   number = 9,
   year = 2018,
   location = "Hyderabad, IN",
   publisher = "International Speech Communication Association",
   ISSN = "1990-9772",
   doi = "10.21437/Interspeech.2018-2361",
   language = "english",
   url = "https://www.fit.vut.cz/research/publication/11844"
}
Back to top