Add to ORKGTo improve speech recognition performance, a combination between TANDEM and bottleneck Deep Neural Networks (DNN) is investigated. In particular, exploiting a feature combination performed by means of a multi-stream hierarchical processing, we show a performance improvement by combining the same input features processed by different neural networks. The experiments are based on the spontaneous telephone recordings of the Cantonese IARPA Babel corpus using both standard MFCCs and Gabor as input features
Effective representation plays an important role in automatic spoken language identification (LID). ...
We propose a new technique for training deep neural networks (DNNs) as data-driven feature front-end...
Our previous work has shown that Deep Bottleneck Features (DBF), generated from a well-trained Deep ...
Bottleneck (BN) feature has attracted considerable attentions by its capacity of improving the accur...
Deep neural networks (DNNs) use a cascade of hidden representa-tions to enable the learning of compl...
In this work, we propose a modular combination of two pop-ular applications of neural networks to la...
Deep neural networks (DNNs) use a cascade of hidden representa-tions to enable the learning of compl...
The recent speaker embeddings framework has been shown to provide excellent performance on the task ...
In this work, a novel training scheme for generating bottleneck fea-tures from deep neural networks ...
In this work, a novel training scheme for generating bottleneck features from deep neural networks i...
Posterior-based or bottleneck features derived from neural net-works trained on out-of-domain data m...
Language recognition systems based on bottleneck features have recently become the state-of-the-art ...
In the tandem approach to modeling the acoustic signal, a neural-net preprocessor is first discrimin...
Recently, the hybrid deep neural network (DNN)-hidden Markov model (HMM) has been shown to significa...
The outputs of multi-layer perceptron (MLP) classifiers have been successfully used in tandem system...
Effective representation plays an important role in automatic spoken language identification (LID). ...
We propose a new technique for training deep neural networks (DNNs) as data-driven feature front-end...
Our previous work has shown that Deep Bottleneck Features (DBF), generated from a well-trained Deep ...
Bottleneck (BN) feature has attracted considerable attentions by its capacity of improving the accur...
Deep neural networks (DNNs) use a cascade of hidden representa-tions to enable the learning of compl...
In this work, we propose a modular combination of two pop-ular applications of neural networks to la...
Deep neural networks (DNNs) use a cascade of hidden representa-tions to enable the learning of compl...
The recent speaker embeddings framework has been shown to provide excellent performance on the task ...
In this work, a novel training scheme for generating bottleneck fea-tures from deep neural networks ...
In this work, a novel training scheme for generating bottleneck features from deep neural networks i...
Posterior-based or bottleneck features derived from neural net-works trained on out-of-domain data m...
Language recognition systems based on bottleneck features have recently become the state-of-the-art ...
In the tandem approach to modeling the acoustic signal, a neural-net preprocessor is first discrimin...
Recently, the hybrid deep neural network (DNN)-hidden Markov model (HMM) has been shown to significa...
The outputs of multi-layer perceptron (MLP) classifiers have been successfully used in tandem system...
Effective representation plays an important role in automatic spoken language identification (LID). ...
We propose a new technique for training deep neural networks (DNNs) as data-driven feature front-end...
Our previous work has shown that Deep Bottleneck Features (DBF), generated from a well-trained Deep ...