Publication detail

Training Data Augmentation and Data Selection

KARAFIÁT, M. VESELÝ, K. ŽMOLÍKOVÁ, K. DELCROIX, M. WATANABE, S. BURGET, L. ČERNOCKÝ, J. SZŐKE, I.

Original Title

Training Data Augmentation and Data Selection

Type

book chapter

Language

English

Original Abstract

Data augmentation is a simple and efficient technique to improve the robustness of a speech recognizer when deployed in mismatched training-test conditions. Our work, conducted during the JSALT 2015 workshop, aimed at the development of: (1) Data augmentation strategies including noising and reverberation. They were tested in combination with two approaches to signal enhancement: a carefully engineered WPE dereverberation and a learned DNN-based denoising autoencoder. (2) Proposing a novel technique for extracting an informative vector from a Sequence Summarizing Neural Network (SSNN). Similarly to i-vector extractor, the SSNN produces a "summary vector", representing an acoustic summary of an utterance. Such vector can be used directly for adaptation, but the main usage matching the aim of this chapter is for selection of augmented training data. All techniques were tested on the AMI training set and CHiME3 test set.

Keywords

training data, augmentation, data selection

Authors

KARAFIÁT, M.; VESELÝ, K.; ŽMOLÍKOVÁ, K.; DELCROIX, M.; WATANABE, S.; BURGET, L.; ČERNOCKÝ, J.; SZŐKE, I.

Released

8. 12. 2017

Publisher

Springer International Publishing

Location

Heidelberg

ISBN

978-3-319-64679-4

Book

New Era for Robust Speech Recognition: Exploiting Deep Learning

Edition

Computer Science, Artificial Intelligence

Pages from

245

Pages to

260

Pages count

16

URL

http://www.springer.com/gp/book/9783319646794#aboutBook

BibTex

@inbook{BUT144497,
  author="Martin {Karafiát} and Karel {Veselý} and Kateřina {Žmolíková} and Marc {Delcroix} and Shinji {Watanabe} and Lukáš {Burget} and Jan {Černocký} and Igor {Szőke}",
  title="Training Data Augmentation and Data Selection",
  booktitle="New Era for Robust Speech Recognition: Exploiting Deep Learning",
  year="2017",
  publisher="Springer International Publishing",
  address="Heidelberg",
  series="Computer Science, Artificial Intelligence",
  pages="245--260",
  doi="10.1007/978-3-319-64680-0\{_}10",
  isbn="978-3-319-64679-4",
  url="http://www.springer.com/gp/book/9783319646794#aboutBook"
}