Improving Lightly Supervised Training for Broadcast TranscriptionReport as inadecuate

Improving Lightly Supervised Training for Broadcast Transcription

Improving Lightly Supervised Training for Broadcast Transcription - Download this document for free, or read online. Document in PDF available to download.

Publication Date: 2013-08-25

Journal Title: Interspeech 2013

Publisher: ISCA

Pages: 2187-2191

Language: English

Type: Article

Metadata: Show full item record

Citation: Long, Y., Gales, M. J. F., Lanchantin, P., Liu, X., Seigel, M. S., & Woodland, P. C. (2013). Improving Lightly Supervised Training for Broadcast Transcription. Interspeech 2013, 2187-2191.

Description: This is the accepted manuscript version. The final version is available at

Abstract: This paper investigates improving lightly supervised acoustic model training for an archive of broadcast data. Standard lightly supervised training uses automatically derived decoding hypotheses using a biased language model. However, as the actual speech can deviate significantly from the original programme scripts that are supplied, the quality of standard lightly supervised hypotheses can be poor. To address this issue, word and segment level combination approaches are used between the lightly supervised transcripts and the original programme scripts which yield improved transcriptions. Experimental results show that systems trained using these improved transcriptions consistently outperform those trained using only the original lightly supervised decoding hypotheses. This is shown to be the case for both the maximum likelihood and minimum phone error trained systems.

Keywords: lightly supervised training, speech recognition, confidence scores

Sponsorship: The research leading to these results was supported by EPSRC Programme Grant EP/I031022/1 (Natural Speech Technology).


This record's URL:

Author: Long, Y.Gales, M. J. F.Lanchantin, P.Liu, X.Seigel, M. S.Woodland, P. C.



Related documents