Massively Multilingual ASR: 50 Languages, 1 Model, 1 Billion Parameters

Vineel Pratap
Vineel Pratap
Anuroop Sriram
Anuroop Sriram
Paden Tomasello
Paden Tomasello
Vitaliy Liptchinsky
Vitaliy Liptchinsky
Gabriel Synnaeve
Gabriel Synnaeve

INTERSPEECH, pp. 4751-4755, 2020.

Cited by: 6|Views38
EI

Abstract:

We study training a single acoustic model for multiple languages with the aim of improving automatic speech recognition (ASR) performance on low-resource languages, and over-all simplifying deployment of ASR systems that support diverse languages. We perform an extensive benchmark on 51 languages, with varying amount of training data by...More

Code:

Data:

Your rating :
0

 

Tags
Comments