A Comparison between Deep Neural Nets and Kernel Acoustic Models for Speech Recognition

03/18/2016
by   Zhiyun Lu, et al.
0

We study large-scale kernel methods for acoustic modeling and compare to DNNs on performance metrics related to both acoustic modeling and recognition. Measuring perplexity and frame-level classification accuracy, kernel-based acoustic models are as effective as their DNN counterparts. However, on token-error-rates DNN models can be significantly better. We have discovered that this might be attributed to DNN's unique strength in reducing both the perplexity and the entropy of the predicted posterior probabilities. Motivated by our findings, we propose a new technique, entropy regularized perplexity, for model selection. This technique can noticeably improve the recognition performance of both types of models, and reduces the gap between them. While effective on Broadcast News, this technique could be also applicable to other tasks.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
01/13/2017

Kernel Approximation Methods for Speech Recognition

We study large-scale kernel methods for acoustic modeling in speech reco...
research
06/30/2014

Building DNN Acoustic Models for Large Vocabulary Speech Recognition

Deep neural networks (DNNs) are now a central component of nearly all st...
research
11/14/2014

How to Scale Up Kernel Methods to Be As Good As Deep Neural Nets

The computational complexity of kernel methods has often been a major ba...
research
08/17/2016

Ensemble of Jointly Trained Deep Neural Network-Based Acoustic Models for Reverberant Speech Recognition

Distant speech recognition is a challenge, particularly due to the corru...
research
11/16/2015

Latent Dirichlet Allocation Based Organisation of Broadcast Media Archives for Deep Neural Network Adaptation

This paper presents a new method for the discovery of latent domains in ...
research
12/21/2018

End-to-End Classification of Reverberant Rooms using DNNs

Reverberation is present in our workplaces, our homes and even in places...
research
06/27/2017

Acoustic Modeling Using a Shallow CNN-HTSVM Architecture

High-accuracy speech recognition is especially challenging when large da...

Please sign up or login with your details

Forgot password? Click here to reset