Topic and keyword identification for low-resourced speech using cross-language transfer learning

Wenda Chen, Mark Allan Hasegawa-Johnson, Nancy F. Chen

Research output: Contribution to journalConference article

Abstract

This paper studies topic and keyword identification for languages in which we have no transcribed speech data. We adopt a transfer learning framework to transfer what is learned from rich-resourced languages (RRL) to low-resourced languages (LRL). Specifically, we propose that a convolutional neural network (CNN) trained as a topic classifier in an RRL learns features (hidden layer activations) that can be used for the same purpose in an LRL. The CNN observes acoustic features, RRL phones, or segment clusters generated by an unsupervised phone clustering system; its hidden layers are retained, and its output layer re-trained from scratch on the LRL. Our results are compared with the state-of-the-art topic classification methods on cross-language ASR transcripts. We also discuss the successful detection of topic dependent keywords and the use of unsupervised learning based clusters in our approach for low-resourced language topic detection.

Original languageEnglish (US)
Pages (from-to)2047-2051
Number of pages5
JournalProceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH
Volume2018-September
DOIs
StatePublished - Jan 1 2018
Event19th Annual Conference of the International Speech Communication, INTERSPEECH 2018 - Hyderabad, India
Duration: Sep 2 2018Sep 6 2018

Fingerprint

Transfer Learning
Neural networks
Unsupervised learning
Classifiers
Acoustics
Chemical activation
Neural Networks
Speech
Language Acquisition
Language
Cross-language Transfer
Key Words
Unsupervised Learning
Activation
Classifier
Clustering

Keywords

  • Low-resourced languages
  • Speech recognition
  • Topic detection

ASJC Scopus subject areas

  • Language and Linguistics
  • Human-Computer Interaction
  • Signal Processing
  • Software
  • Modeling and Simulation

Cite this

@article{8d4b5a28b8964cb8bba15512dcad5f3e,
title = "Topic and keyword identification for low-resourced speech using cross-language transfer learning",
abstract = "This paper studies topic and keyword identification for languages in which we have no transcribed speech data. We adopt a transfer learning framework to transfer what is learned from rich-resourced languages (RRL) to low-resourced languages (LRL). Specifically, we propose that a convolutional neural network (CNN) trained as a topic classifier in an RRL learns features (hidden layer activations) that can be used for the same purpose in an LRL. The CNN observes acoustic features, RRL phones, or segment clusters generated by an unsupervised phone clustering system; its hidden layers are retained, and its output layer re-trained from scratch on the LRL. Our results are compared with the state-of-the-art topic classification methods on cross-language ASR transcripts. We also discuss the successful detection of topic dependent keywords and the use of unsupervised learning based clusters in our approach for low-resourced language topic detection.",
keywords = "Low-resourced languages, Speech recognition, Topic detection",
author = "Wenda Chen and Hasegawa-Johnson, {Mark Allan} and Chen, {Nancy F.}",
year = "2018",
month = "1",
day = "1",
doi = "10.21437/Interspeech.2018-1283",
language = "English (US)",
volume = "2018-September",
pages = "2047--2051",
journal = "Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH",
issn = "2308-457X",

}

TY - JOUR

T1 - Topic and keyword identification for low-resourced speech using cross-language transfer learning

AU - Chen, Wenda

AU - Hasegawa-Johnson, Mark Allan

AU - Chen, Nancy F.

PY - 2018/1/1

Y1 - 2018/1/1

N2 - This paper studies topic and keyword identification for languages in which we have no transcribed speech data. We adopt a transfer learning framework to transfer what is learned from rich-resourced languages (RRL) to low-resourced languages (LRL). Specifically, we propose that a convolutional neural network (CNN) trained as a topic classifier in an RRL learns features (hidden layer activations) that can be used for the same purpose in an LRL. The CNN observes acoustic features, RRL phones, or segment clusters generated by an unsupervised phone clustering system; its hidden layers are retained, and its output layer re-trained from scratch on the LRL. Our results are compared with the state-of-the-art topic classification methods on cross-language ASR transcripts. We also discuss the successful detection of topic dependent keywords and the use of unsupervised learning based clusters in our approach for low-resourced language topic detection.

AB - This paper studies topic and keyword identification for languages in which we have no transcribed speech data. We adopt a transfer learning framework to transfer what is learned from rich-resourced languages (RRL) to low-resourced languages (LRL). Specifically, we propose that a convolutional neural network (CNN) trained as a topic classifier in an RRL learns features (hidden layer activations) that can be used for the same purpose in an LRL. The CNN observes acoustic features, RRL phones, or segment clusters generated by an unsupervised phone clustering system; its hidden layers are retained, and its output layer re-trained from scratch on the LRL. Our results are compared with the state-of-the-art topic classification methods on cross-language ASR transcripts. We also discuss the successful detection of topic dependent keywords and the use of unsupervised learning based clusters in our approach for low-resourced language topic detection.

KW - Low-resourced languages

KW - Speech recognition

KW - Topic detection

UR - http://www.scopus.com/inward/record.url?scp=85054976924&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85054976924&partnerID=8YFLogxK

U2 - 10.21437/Interspeech.2018-1283

DO - 10.21437/Interspeech.2018-1283

M3 - Conference article

VL - 2018-September

SP - 2047

EP - 2051

JO - Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

JF - Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

SN - 2308-457X

ER -