Speech-based diagnosis of autism spectrum condition by generative adversarial network representations

Jun Deng, Nicholas Cummins, Maximilian Schmitt, Kun Qian, Fabien Ringeval, Björn Schuller

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

33 Scopus citations

Abstract

Machine learning paradigms based on child vocalisations show great promise as an objective marker of developmental disorders such as Autism. In conventional detection systems, hand-crafted acoustic features are usually fed into a discriminative classifier (e. g., Support Vector Machines); however it is well known that the accuracy and robustness of such a system is limited by the size of the associated training data. This paper explores, for the first time, the use of feature representations learnt using a deep Generative Adversarial Network (GAN) for classifying children's speech affected by developmental disorders. A comparative evaluation of our proposed system with different acoustic feature sets is performed on the Child Pathological and Emotional Speech database. Key experimental results presented demonstrate that GAN based methods exhibit competitive performance with the conventional paradigms in terms of the unweighted average recall metric.

Original languageEnglish
Title of host publicationDH 2017 - Proceedings of the 2017 International Conference on Digital Health
PublisherAssociation for Computing Machinery
Pages53-57
Number of pages5
ISBN (Electronic)9781450352499
DOIs
StatePublished - 2 Jul 2017
Externally publishedYes
Event7th International Conference on Digital Health, DH 2017 - London, United Kingdom
Duration: 2 Jul 20175 Jul 2017

Publication series

NameACM International Conference Proceeding Series
VolumePart F128634

Conference

Conference7th International Conference on Digital Health, DH 2017
Country/TerritoryUnited Kingdom
CityLondon
Period2/07/175/07/17

Keywords

  • Autism spectrum condition
  • Automatic diagnosis
  • Generative adversarial networks
  • Representation learning

Fingerprint

Dive into the research topics of 'Speech-based diagnosis of autism spectrum condition by generative adversarial network representations'. Together they form a unique fingerprint.

Cite this