[ANGLÈS] This work focuses on building and testing statistical language models based on recurrent neural networks. Although both speaker independent and dependent language models will be discussed and compared, this thesis makes emphasis on the speaker dependent scenario when building complete automatic speech recognition systems. Traditional techniques for estimating language models are based on N-gram counts and they basically remained the state-of-the-art for many applications. Recently, faster CPUs and efficient techniques made it possible to apply RNN-based language models to state-of-the-art systems efficiently. In this Thesis, you will see reductions in perplexity of RNN-based speaker dependent language models up to 25\% relative. Th...
Comunicació presentada a la 2016 Conference of the North American Chapter of the Association for Com...
The task of part-of-speech (POS) language modeling typically includes a very small vocabulary, which...
In recent years, recurrent neural network language models (RNNLMs) have become increasingly popular ...
[ANGLÈS] This work focuses on building and testing statistical language models based on recurrent ne...
Deep learning has revolutionized almost every engineering branch over the past decades and have also...
A language model based in continuous representations of words is presented, which has been applied t...
In this paper we present a survey on the application of recurrent neural networks to the task of sta...
AbstractIn this paper, we present a survey on the application of recurrent neural networks to the ta...
The goal of this thesis is to advance the use of recurrent neural network language models (RNNLMs) ...
Le domaine du traitement automatique de la parole regroupe un très grand nombre de tâches parmi lesq...
Virtually any modern speech recognition system relies on count-based language models. In this thesis...
Recurrent neural network language models (RNNLMs) are powerful language modeling techniques. Signifi...
ABSTRACT We present several modifications of the original recurrent neural network language model (R...
Recurrent neural network language models (RNNLMs) have recently shown to outperform the venerable n-...
This thesis focuses on the automatic construction of linguistic tools and resources for analyzing te...
Comunicació presentada a la 2016 Conference of the North American Chapter of the Association for Com...
The task of part-of-speech (POS) language modeling typically includes a very small vocabulary, which...
In recent years, recurrent neural network language models (RNNLMs) have become increasingly popular ...
[ANGLÈS] This work focuses on building and testing statistical language models based on recurrent ne...
Deep learning has revolutionized almost every engineering branch over the past decades and have also...
A language model based in continuous representations of words is presented, which has been applied t...
In this paper we present a survey on the application of recurrent neural networks to the task of sta...
AbstractIn this paper, we present a survey on the application of recurrent neural networks to the ta...
The goal of this thesis is to advance the use of recurrent neural network language models (RNNLMs) ...
Le domaine du traitement automatique de la parole regroupe un très grand nombre de tâches parmi lesq...
Virtually any modern speech recognition system relies on count-based language models. In this thesis...
Recurrent neural network language models (RNNLMs) are powerful language modeling techniques. Signifi...
ABSTRACT We present several modifications of the original recurrent neural network language model (R...
Recurrent neural network language models (RNNLMs) have recently shown to outperform the venerable n-...
This thesis focuses on the automatic construction of linguistic tools and resources for analyzing te...
Comunicació presentada a la 2016 Conference of the North American Chapter of the Association for Com...
The task of part-of-speech (POS) language modeling typically includes a very small vocabulary, which...
In recent years, recurrent neural network language models (RNNLMs) have become increasingly popular ...