Loading [MathJax]/extensions/TeX/ietmacros.js
Language Model Bootstrapping Using Neural Machine Translation for Conversational Speech Recognition | IEEE Conference Publication | IEEE Xplore

Language Model Bootstrapping Using Neural Machine Translation for Conversational Speech Recognition


Abstract:

Building conversational speech recognition systems for new languages is constrained by the availability of utterances capturing user-device interactions. Data collection ...Show More

Abstract:

Building conversational speech recognition systems for new languages is constrained by the availability of utterances capturing user-device interactions. Data collection is expensive and limited by speed of manual transcription. In order to address this, we advocate the use of neural machine translation as a data augmentation technique for bootstrapping language models. Machine translation (MT) offers a systematic way of incorporating collections from mature, resource-rich conversational systems that may be available for a different language. However, ingesting raw translations from a general purpose MT system may not be effective owing to the presence of named entities, intra sentential code-switching and the domain mismatch between the conversational data being translated and the parallel text used for MT training. To circumvent this, we explore following domain adaptation techniques: (a) sentence embedding based data selection for MT training, (b) model finetuning, and (c) rescoring and filtering translated hypotheses. Using Hindi language as the experimental testbed, we supplement transcribed collections with translated US English utterances. We observe a relative word error rate reduction of 7.8-15.6%, depending on the bootstrapping phase. Fine grained analysis reveals that translation particularly aids the interaction scenarios underrepresented in the transcribed data.
Date of Conference: 14-18 December 2019
Date Added to IEEE Xplore: 20 February 2020
ISBN Information:
Conference Location: Singapore

References

References is not available for this document.