|
SraVaani ASR Model Trained on 65 Indian Languages and Dialects
Researchers at IISc’s SPIRE Lab, in collaboration with ARTPARK and Google, release the SraVaani model.
SraVaani is a multilingual Automatic Speech Recognition (ASR) model trained on 65 Indian languages and dialects.
SraVaani Multilingual ASR Model:
| Dimension | Key Details |
|---|---|
| Developers | SraVaani is released by researchers at IISc’s SPIRE Lab in collaboration with ARTPARK and Google. |
| Model type | SraVaani comprises a multilingual model for Automatic Speech Recognition (ASR) systems. |
| Training coverage | SraVaani is trained on 65 Indian languages and dialects. |
| Language composition | SraVaani covers 20 scheduled languages, and 45 regional languages and dialects. |
| Dataset base | SraVaani is built on top of the Vaani dataset, one of the largest speech corpora for Indian languages. |