Modelling out-of-vocabulary words for robust speech recognition
By Issam Bazzi, James R. Glass. 2002
205 citations recorded by the source corpus.
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
- Li Deng
- Dong Yu
- Sabato Marco Siniscalchi
- Yoshua Bengio
- Jinyu Li
- Chin-Hui Lee
- James R. Glass
- Hlib Babii
Citations in the corpus
- Big Code != Big Vocabulary: Open-Vocabulary Models for Source Code (2020, methodology): “ers that occursas a subsequence ofsometoken inthetraining set;the model outputs a sequence of subword units instead of a sequence of tokens. Many NLP models have used linguistically-motivated subwords[11,24,59,65].Mikolovetal.foundthatsubwordmodels improved on character models [65].Sennrich et al. a”
- Maybe Deep Neural Networks are the Best Choice for Modeling Source Code (2019, background): “For example, language models have been proposed at the morpheme level[49], the phone level [10], and the syllable level [55].”
- Automatic alignment and error correction of human generated transcripts for long speech recordings (2006, methodology): “Instead, we assume that errors in the transcript are possible and we allow insertions of new words and substitutions for existing words through the use of a phonetic-based out-of-vocabulary (OOV) word filler model [ 3 ].”
- Exploiting deep neural networks for detection-based speech recognition (2013, background): “Finally, phoneme estimation also plays a very important role in many speech processing applications, such as out-ofvocabulary detection [41] and language identification (e.”
- A Survey on Text Classification: From Traditional to Deep Learning (2020, unknown): “Character-level embeddings can settle Out-OfVocabulary (OOV) [185] words.”
- Discovering linguistic structures in speech: models and applications (2014, methodology): “Similar to our approach, the OOV detectors represent words as sub-word sequences, which are shared and re-used across spoken words in a language; however, the sub-word units employed in most OOV detectors are induced from a pre-defined phonetic unit set [6, 145].”