Massively Multilingual ASR: 50 Languages, 1 Model, 1 Billion Parameters
| Autor: | Ronan Collobert, Gabriel Synnaeve, Paden Tomasello, Vitaliy Liptchinsky, Awni Hannun, Vineel Pratap, Anuroop Sriram |
|---|---|
| Jazyk: | angličtina |
| Rok vydání: | 2020 |
| Předmět: |
FOS: Computer and information sciences
Sound (cs.SD) Training set Computer Science - Computation and Language Computer science business.industry Low resource Acoustic model computer.software_genre Input language Computer Science - Sound Reduction (complexity) Audio and Speech Processing (eess.AS) Head model Benchmark (computing) FOS: Electrical engineering electronic engineering information engineering Artificial intelligence Joint (audio engineering) business computer Computation and Language (cs.CL) Natural language processing Electrical Engineering and Systems Science - Audio and Speech Processing |
| Zdroj: | INTERSPEECH |
| Popis: | 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 language(from 100 hours to 1100 hours). We compare three variants of multilingual training from a single joint model without knowing the input language, to using this information, to multiple heads (one per language cluster). We show that multilingual training of ASR models on several languages can improve recognition performance, in particular, on low resource languages. We see 20.9%, 23% and 28.8% average WER relative reduction compared to monolingual baselines on joint model, joint model with language input and multi head model respectively. To our knowledge, this is the first work studying multilingual ASR at massive scale, with more than 50 languages and more than 16,000 hours of audio across them. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|