Low-frequency words place a major challenge for automatic speech recognition (ASR). The probabilities of these words, which are often important name entities, are generally under-estimated by the language model (LM) due to their limited occurrences in the training data. Recently, we proposed a word-pair approach to deal with the problem, which borrows information of frequent words to enhance the probabilities of low-frequency words. This paper presents an extension to the word-pair method by involving multiple `predicting words' to produce better estimation for low-frequency words. We also employ this approach to deal with out-of-language words in the task of multi-lingual speech recognition.
Cite as: Ma, X., Wang, X., Wang, D., Zhang, Z. (2015) Recognize foreign low-frequency words with similar pairs. Proc. Interspeech 2015, 458-462, doi: 10.21437/Interspeech.2015-175
@inproceedings{ma15b_interspeech,
author={Xi Ma and Xiaoxi Wang and Dong Wang and Zhiyong Zhang},
title={{Recognize foreign low-frequency words with similar pairs}},
year=2015,
booktitle={Proc. Interspeech 2015},
pages={458--462},
doi={10.21437/Interspeech.2015-175}
}