Exploiting Convolutional Neural Networks for Phonotactic Based Dialect Identification

Maryam Najafian; Sameer Khurana; Suwon Shon; Ahmed Ali; James R. Glass

ICASSP 2018accepted0 citations

Exploiting Convolutional Neural Networks for Phonotactic Based Dialect Identification

Maryam Najafian, Sameer Khurana, Suwon Shon, Ahmed Ali, James R. Glass

Abstract

In this paper, we investigate different approaches for Dialect Identification (DID) in Arabic broadcast speech. Dialects differ in their inventory of phonological segments. This paper proposes a new phonotactic based feature representation approach which enables discrimination among different occurrences of the same phone n-grams with different phone duration and probability statistics. To achieve further gain in accuracy we used multi-lingual phone recognizers, trained separately on Arabic, English, Czech, Hungarian and Russian languages. We use Support Vector Machines (SVMs), and Convolutional Neural Networks (CNN s) as backend classifiers throughout the study. The final system fusion results in 24.7% and 19.0% relative error rate reduction compared to that of a conventional phonotactic DID, and i-vectors with bottleneck features.

BibTeX

@inproceedings{icassp2018_exploitingconvol,
  title = {Exploiting Convolutional Neural Networks for Phonotactic Based Dialect Identification},
  author = {Maryam Najafian and Sameer Khurana and Suwon Shon and Ahmed Ali and James R. Glass},
  booktitle = {ICASSP 2018},
  year = {2018}
}