Large vocabulary Russian speech recognition using syntactico-statistical language modeling

Karpov, Alexey; Markov, Konstantin; Kipyatkova, Irina; Vazhenina, Daria; Ronzhin, Andrey

doi:10.1016/j.specom.2013.07.004

Cited by 51 publications

(28 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…2. The system works in 2 modes [15]: training and recognition. In the training mode, acoustic models of speech units, LMs, and phonemic vocabulary of word-forms that will be used by recognizer are created.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Language Models with RNNs for Rescoring Hypotheses of Russian ASR

Kipyatkova

Karpov

2016

Advances in Neural Networks – ISNN 2016

Self Cite

View full text Add to dashboard Cite

Abstract. In this paper, we describe a research of recurrent neural networks (RNNs) for language modeling in large vocabulary continuous speech recognition for Russian. We experimented with recurrent neural networks with different number of units in the hidden layer. RNN-based and 3-gram language models (LMs) were trained using the text corpus of 350M words. Obtained RNN-based language models were used for N-best list rescoring for automatic continuous Russian speech recognition. We tested also a linear interpolation of RNN LMs with the baseline 3-gram LM and achieved 14 % relative reduction of the word error rate (WER) with respect to the baseline 3-gram model.

show abstract

Section: Methodsmentioning

confidence: 99%

“…The procedure of preliminary text processing and normalization is described in [15]. At first, texts were divided into sentences.…”

Section: Training Textual Corpus and Baseline Language Modelmentioning

confidence: 99%

Language Models with RNNs for Rescoring Hypotheses of Russian ASR

Kipyatkova

Karpov

2016

Advances in Neural Networks – ISNN 2016

Self Cite

View full text Add to dashboard Cite

show abstract

“…For instance, Russian language [13,34] is rather difficult due to the following reasons: (1) of word forms; (2) the number of vowels (5)(6)in Russian is much lower in comparison with 17 vowels with a lot of diphthongs in English, hence, the phoneme recognition accuracy is low. Nevertheless, the state-of-the-art approach to ASR can be applied even in Russian by using the multilingual training of acoustic model [12].…”

Section: Introductionmentioning

confidence: 99%

Towards the creation of reliable voice control system based on a fuzzy approach

Savchenko

2015

Pattern Recognition Letters

View full text Add to dashboard Cite

“…It should be noted that efficiency of ngram models is heavily language dependent. They correspond well to grammatical structure of positional languages (such as English), but in case of Polish and other highly inflected languages, words order is not a key indicator of relations between them [8]. The main difficulty in language modelling and learning problems in general is the curse of dimensionality.…”

mentioning

confidence: 99%

Comparison of Language Models Trained on Written Texts and Speech Transcripts in the Context of Automatic Speech Recognition

Dziadzio

Nabozny

Smywiński-Pohl

et al. 2015

Annals of Computer Science and Information Systems

View full text Add to dashboard Cite

Abstract-We investigate whether language models used in automatic speech recognition (ASR) should be trained on speech transcripts rather than on written texts. By calculating log-likelihood statistic for part-of-speech (POS) n-grams, we show that there are significant differences between written texts and speech transcripts. We also test the performance of language models trained on speech transcripts and written texts in ASR and show that using the former results in greater word error reduction rates (WERR), even if the model is trained on much smaller corpora. For our experiments we used the manually labeled one million subcorpus of the National Corpus of Polish and an HTK acoustic model. Index Terms-automatic speech recognition, morphosyntactic language model, written and spoken language comparison

show abstract

Large vocabulary Russian speech recognition using syntactico-statistical language modeling

Cited by 51 publications

References 23 publications

Language Models with RNNs for Rescoring Hypotheses of Russian ASR

Language Models with RNNs for Rescoring Hypotheses of Russian ASR

Towards the creation of reliable voice control system based on a fuzzy approach

Comparison of Language Models Trained on Written Texts and Speech Transcripts in the Context of Automatic Speech Recognition

Contact Info

Product

Resources

About