Publication Details

A multi purpose and large scale speech corpus in Persian and English for speaker and speech Recognition: the DeepMine database

ZEINALI Hossein, ČERNOCKÝ Jan and BURGET Lukáš. A multi purpose and large scale speech corpus in Persian and English for speaker and speech Recognition: the DeepMine database. In: Proceedings of ASRU 2019. Sentosa, Singapore: IEEE Signal Processing Society, 2019, pp. 397-402. ISBN 978-1-72810-306-8.
Czech title
Víceúčelový velký řečový korpus perštiny a angličtiny pro rozpoznávání řeči a řečníka: databáze DeepMine
Type
conference paper
Language
english
Authors
URL
Keywords
speech database, text-dependent, text-independent, speaker verification, speech recognition
Abstract
DeepMine is a speech database in Persian and English designed to build and evaluate text-dependent, text-prompted, and textindependent speaker verification, as well as Persian speech recognition systems. It contains more than 1850 speakers and 540 thousand recordings overall, more than 480 hours of speech are transcribed. It is the first public large-scale speaker verification database in Persian, the largest public text-dependent and text-prompted speaker verification database in English, and the largest public evaluation dataset for text-independent speaker verification. It has a good coverage of age, gender, and accents. We provide several evaluation protocols for each part of the database to allow for research on different aspects of speaker verification. We also provide the results of several experiments that can be considered as baselines: HMM-based i-vectors for text-dependent speaker verification, and HMM-based as well as state-of-the-art deep neural network based ASR. We demonstrate that the database can serve for training robust ASR models.
Published
2019
Pages
397-402
Proceedings
Proceedings of ASRU 2019
Conference
Automatic Speech Recognition and Understanding Workshop 2019, Automatic Speech Recognition and Understanding Workshop 2019, SG
ISBN
978-1-72810-306-8
Publisher
IEEE Signal Processing Society
Place
Sentosa, Singapore, SG
BibTeX
@INPROCEEDINGS{FITPUB12153,
   author = "Hossein Zeinali and Jan \v{C}ernock\'{y} and Luk\'{a}\v{s} Burget",
   title = "A multi purpose and large scale speech corpus in Persian and English for speaker and speech Recognition: the DeepMine database",
   pages = "397--402",
   booktitle = "Proceedings of ASRU 2019",
   year = 2019,
   location = "Sentosa, Singapore, SG",
   publisher = "IEEE Signal Processing Society",
   ISBN = "978-1-72810-306-8",
   language = "english",
   url = "https://www.fit.vut.cz/research/publication/12153"
}
Back to top