Improving Sequence-To-Sequence Speech Recognition Training with On-The-Fly Data Augmentation

Sequence-to-Sequence (S2S) models recently started to show state-of-the-art performance for automatic speech recognition (ASR). With these large and deep models overfitting remains the largest problem, outweighing performance improvements that can be obtained from better architectures. One solution...

Full description

Saved in:
Bibliographic Details
Main Authors: Nguyen, Thai-Son, Stuker, Sebastian, Niehues, Jan, Waibel, Alex
Format: Conference Proceeding
Language:English
Subjects:
Data Augmentation
Data models
Perturbation methods
Self-attention
Sequence-to-sequence
Signal processing
Speech processing
Speed Perturbation
Sub-sequence
Switches
Training
Training data
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Sequence-to-Sequence (S2S) models recently started to show state-of-the-art performance for automatic speech recognition (ASR). With these large and deep models overfitting remains the largest problem, outweighing performance improvements that can be obtained from better architectures. One solution to the overfitting problem is increasing the amount of available training data and the variety exhibited by the training data with the help of data augmentation. In this paper we examine the influence of three data augmentation methods on the performance of two S2S model architectures. One of the data augmentation method comes from literature, while two other methods are our own development - a time perturbation in the frequency domain and sub-sequence sampling. Our experiments on Switchboard and Fisher data show state-of-the-art performance for S2S models that are trained solely on the speech training data and do not use additional text data.
ISSN:2379-190X
DOI:10.1109/ICASSP40776.2020.9054130