The MGB-2 challenge

Arabic multi-dialect broadcast media recognition

Ahmed Ali, Peter Bell, James Glass, Yacine Messaoui, Hamdy Mubarak, Steve Renals, Yifan Zhang

Research output: Chapter in Book/Report/Conference proceedingConference contribution

15 Citations (Scopus)

Abstract

This paper describes the Arabic Multi-Genre Broadcast (MGB-2) Challenge for SLT-2016. Unlike last year's English MGB Challenge, which focused on recognition of diverse TV genres, this year, the challenge has an emphasis on handling the diversity in dialect in Arabic speech. Audio data comes from 19 distinct programmes from the Aljazeera Arabic TV channel between March 2005 and December 2015. Programmes are split into three groups: conversations, interviews, and reports. A total of 1,200 hours have been released with lightly supervised transcriptions for the acoustic modelling. For language modelling, we made available over 110M words crawled from Aljazeera Arabic website Aljazeera.net for a 10 year duration 2000-2011. Two lexicons have been provided, one phoneme based and one grapheme based. Finally, two tasks were proposed for this year's challenge: standard speech transcription, and word alignment. This paper describes the task data and evaluation process used in the MGB challenge, and summarises the results obtained.

Original languageEnglish
Title of host publication2016 IEEE Workshop on Spoken Language Technology, SLT 2016 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages279-284
Number of pages6
ISBN (Electronic)9781509049035
DOIs
Publication statusPublished - 7 Feb 2017
Event2016 IEEE Workshop on Spoken Language Technology, SLT 2016 - San Diego, United States
Duration: 13 Dec 201616 Dec 2016

Other

Other2016 IEEE Workshop on Spoken Language Technology, SLT 2016
CountryUnited States
CitySan Diego
Period13/12/1616/12/16

Fingerprint

Transcription
Websites
Acoustics

Keywords

  • Alignment
  • Broadcast speech
  • Multi-genre
  • Speech recognition
  • Transcription

ASJC Scopus subject areas

  • Human-Computer Interaction
  • Artificial Intelligence
  • Language and Linguistics
  • Computer Vision and Pattern Recognition
  • Computer Science Applications

Cite this

Ali, A., Bell, P., Glass, J., Messaoui, Y., Mubarak, H., Renals, S., & Zhang, Y. (2017). The MGB-2 challenge: Arabic multi-dialect broadcast media recognition. In 2016 IEEE Workshop on Spoken Language Technology, SLT 2016 - Proceedings (pp. 279-284). [7846277] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SLT.2016.7846277

The MGB-2 challenge : Arabic multi-dialect broadcast media recognition. / Ali, Ahmed; Bell, Peter; Glass, James; Messaoui, Yacine; Mubarak, Hamdy; Renals, Steve; Zhang, Yifan.

2016 IEEE Workshop on Spoken Language Technology, SLT 2016 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2017. p. 279-284 7846277.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Ali, A, Bell, P, Glass, J, Messaoui, Y, Mubarak, H, Renals, S & Zhang, Y 2017, The MGB-2 challenge: Arabic multi-dialect broadcast media recognition. in 2016 IEEE Workshop on Spoken Language Technology, SLT 2016 - Proceedings., 7846277, Institute of Electrical and Electronics Engineers Inc., pp. 279-284, 2016 IEEE Workshop on Spoken Language Technology, SLT 2016, San Diego, United States, 13/12/16. https://doi.org/10.1109/SLT.2016.7846277
Ali A, Bell P, Glass J, Messaoui Y, Mubarak H, Renals S et al. The MGB-2 challenge: Arabic multi-dialect broadcast media recognition. In 2016 IEEE Workshop on Spoken Language Technology, SLT 2016 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2017. p. 279-284. 7846277 https://doi.org/10.1109/SLT.2016.7846277
Ali, Ahmed ; Bell, Peter ; Glass, James ; Messaoui, Yacine ; Mubarak, Hamdy ; Renals, Steve ; Zhang, Yifan. / The MGB-2 challenge : Arabic multi-dialect broadcast media recognition. 2016 IEEE Workshop on Spoken Language Technology, SLT 2016 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2017. pp. 279-284
@inproceedings{97ac0161df2242148937fb2d0b7a9ef3,
title = "The MGB-2 challenge: Arabic multi-dialect broadcast media recognition",
abstract = "This paper describes the Arabic Multi-Genre Broadcast (MGB-2) Challenge for SLT-2016. Unlike last year's English MGB Challenge, which focused on recognition of diverse TV genres, this year, the challenge has an emphasis on handling the diversity in dialect in Arabic speech. Audio data comes from 19 distinct programmes from the Aljazeera Arabic TV channel between March 2005 and December 2015. Programmes are split into three groups: conversations, interviews, and reports. A total of 1,200 hours have been released with lightly supervised transcriptions for the acoustic modelling. For language modelling, we made available over 110M words crawled from Aljazeera Arabic website Aljazeera.net for a 10 year duration 2000-2011. Two lexicons have been provided, one phoneme based and one grapheme based. Finally, two tasks were proposed for this year's challenge: standard speech transcription, and word alignment. This paper describes the task data and evaluation process used in the MGB challenge, and summarises the results obtained.",
keywords = "Alignment, Broadcast speech, Multi-genre, Speech recognition, Transcription",
author = "Ahmed Ali and Peter Bell and James Glass and Yacine Messaoui and Hamdy Mubarak and Steve Renals and Yifan Zhang",
year = "2017",
month = "2",
day = "7",
doi = "10.1109/SLT.2016.7846277",
language = "English",
pages = "279--284",
booktitle = "2016 IEEE Workshop on Spoken Language Technology, SLT 2016 - Proceedings",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - The MGB-2 challenge

T2 - Arabic multi-dialect broadcast media recognition

AU - Ali, Ahmed

AU - Bell, Peter

AU - Glass, James

AU - Messaoui, Yacine

AU - Mubarak, Hamdy

AU - Renals, Steve

AU - Zhang, Yifan

PY - 2017/2/7

Y1 - 2017/2/7

N2 - This paper describes the Arabic Multi-Genre Broadcast (MGB-2) Challenge for SLT-2016. Unlike last year's English MGB Challenge, which focused on recognition of diverse TV genres, this year, the challenge has an emphasis on handling the diversity in dialect in Arabic speech. Audio data comes from 19 distinct programmes from the Aljazeera Arabic TV channel between March 2005 and December 2015. Programmes are split into three groups: conversations, interviews, and reports. A total of 1,200 hours have been released with lightly supervised transcriptions for the acoustic modelling. For language modelling, we made available over 110M words crawled from Aljazeera Arabic website Aljazeera.net for a 10 year duration 2000-2011. Two lexicons have been provided, one phoneme based and one grapheme based. Finally, two tasks were proposed for this year's challenge: standard speech transcription, and word alignment. This paper describes the task data and evaluation process used in the MGB challenge, and summarises the results obtained.

AB - This paper describes the Arabic Multi-Genre Broadcast (MGB-2) Challenge for SLT-2016. Unlike last year's English MGB Challenge, which focused on recognition of diverse TV genres, this year, the challenge has an emphasis on handling the diversity in dialect in Arabic speech. Audio data comes from 19 distinct programmes from the Aljazeera Arabic TV channel between March 2005 and December 2015. Programmes are split into three groups: conversations, interviews, and reports. A total of 1,200 hours have been released with lightly supervised transcriptions for the acoustic modelling. For language modelling, we made available over 110M words crawled from Aljazeera Arabic website Aljazeera.net for a 10 year duration 2000-2011. Two lexicons have been provided, one phoneme based and one grapheme based. Finally, two tasks were proposed for this year's challenge: standard speech transcription, and word alignment. This paper describes the task data and evaluation process used in the MGB challenge, and summarises the results obtained.

KW - Alignment

KW - Broadcast speech

KW - Multi-genre

KW - Speech recognition

KW - Transcription

UR - http://www.scopus.com/inward/record.url?scp=85016058086&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85016058086&partnerID=8YFLogxK

U2 - 10.1109/SLT.2016.7846277

DO - 10.1109/SLT.2016.7846277

M3 - Conference contribution

SP - 279

EP - 284

BT - 2016 IEEE Workshop on Spoken Language Technology, SLT 2016 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

ER -