CUED Publications database

The MGB challenge: Evaluating multi-genre broadcast media recognition

Bell, P and Gales, MJF and Hain, T and Kilgour, J and Lanchantin, P and Liu, X and McParland, A and Renals, S and Saz, O and Wester, M and Woodland, PC (2016) The MGB challenge: Evaluating multi-genre broadcast media recognition. In: UNSPECIFIED pp. 687-693..

Full text not available from this repository.


© 2015 IEEE. This paper describes the Multi-Genre Broadcast (MGB) Challenge at ASRU 2015, an evaluation focused on speech recognition, speaker diarization, and «lightly supervised» alignment of BBC TV recordings. The challenge training data covered the whole range of seven weeks BBC TV output across four channels, resulting in about 1,600 hours of broadcast audio. In addition several hundred million words of BBC subtitle text was provided for language modelling. A novel aspect of the evaluation was the exploration of speech recognition and speaker diarization in a longitudinal setting - i.e. recognition of several episodes of the same show, and speaker diarization across these episodes, linking speakers. The longitudinal tasks also offered the opportunity for systems to make use of supplied metadata including show title, genre tag, and date/time of transmission. This paper describes the task data and evaluation process used in the MGB challenge, and summarises the results obtained.

Item Type: Conference or Workshop Item (UNSPECIFIED)
Divisions: Div F > Machine Intelligence
Depositing User: Cron Job
Date Deposited: 17 Jul 2017 19:01
Last Modified: 21 Jun 2018 02:26