Very deep multilingual convolutional neural networks for LVCSR

T Sercu, C Puhrsch, B Kingsbury… - 2016 IEEE international …, 2016‏ - ieeexplore.ieee.org
Convolutional neural networks (CNNs) are a standard component of many current state-of-
the-art Large Vocabulary Continuous Speech Recognition (LVCSR) systems. However …

[PDF][PDF] Recurrent Neural Network Based Language Modeling in Meeting Recognition.

S Kombrink, T Mikolov, M Karafiát, L Burget - Interspeech, 2011‏ - academia.edu
We use recurrent neural network (RNN) based language models to improve the BUT
English meeting recognizer. On the baseline setup using the original language models we …

[PDF][PDF] Neural Network Bottleneck Features for Language Identification.

P Matejka, Le Zhang 0002, T Ng, O Glembek, JZ Ma… - Odyssey, 2014‏ - isca-archive.org
This paper presents the application of Neural Network Bottleneck (BN) features in Language
Identification (LID). BN features are generally used for Large Vocabulary Speech …

Multilingual representations for low resource speech recognition and keyword search

J Cui, B Kingsbury, B Ramabhadran… - 2015 IEEE workshop …, 2015‏ - ieeexplore.ieee.org
This paper examines the impact of multilingual (ML) acoustic representations on Automatic
Speech Recognition (ASR) and keyword search (KWS) for low resource languages in the …

Multilingually trained bottleneck features in spoken language recognition

R Fer, P Matějka, F Grézl, O Plchot, K Veselý… - Computer Speech & …, 2017‏ - Elsevier
Multilingual training of neural networks has proven to be simple yet effective way to deal with
multilingual training corpora. It allows to use several resources to jointly train a language …

[HTML][HTML] On the use of deep feedforward neural networks for automatic language identification

I Lopez-Moreno, J Gonzalez-Dominguez… - Computer Speech & …, 2016‏ - Elsevier
In this work, we present a comprehensive study on the use of deep neural networks (DNNs)
for automatic language identification (LID). Motivated by the recent success of using DNNs …

Transcribing meetings with the AMIDA systems

T Hain, L Burget, J Dines, PN Garner… - … on Audio, Speech …, 2011‏ - ieeexplore.ieee.org
In this paper, we give an overview of the AMIDA systems for transcription of conference and
lecture room meetings. The systems were developed for participation in the Rich …

Adaptation of multilingual stacked bottle-neck neural network structure for new language

F Grézl, M Karafiát, K Veselý - 2014 IEEE International …, 2014‏ - ieeexplore.ieee.org
The neural network based features became an inseparable part of state-of-the-art LVCSR
systems. In order to perform well, the network has to be trained on a large amount of in …

HMM-based phrase-independent i-vector extractor for text-dependent speaker verification

H Zeinali, H Sameti, L Burget - IEEE/ACM Transactions on …, 2017‏ - ieeexplore.ieee.org
The low-dimensional i-vector representation of speech segments is used in the state-of-the-
art text-independent speaker verification systems. However, i-vectors were deemed …

Using neural network front-ends on far field multiple microphones based speech recognition

Y Liu, P Zhang, T Hain - 2014 IEEE international conference on …, 2014‏ - ieeexplore.ieee.org
This paper presents an investigation of far field speech recognition using beamforming and
channel concatenation in the context of Deep Neural Network (DNN) based feature …