SraVaani
Prelims:
Science &
Technology
Mains:
GS Paper II: Issues
relating to development and inclusion, GS Paper III: Developments and
applications of Science & Technology
Current relevance:
Researchers at the SPIRE Lab
of the Indian Institute of Science (IISc), Bengaluru, in
collaboration with ARTPARK and Google, have released SraVaani.
Highlights:
1.
About SraVaani:
i.
SraVaani is a multilingual Indian speech-recognition
model trained on 65 Indian languages and dialects.
ii.
It has been developed by
researchers at IISc's SPIRE Lab, in collaboration with ARTPARK and
Google.
iii.
It extends Automatic
Speech Recognition (ASR) beyond widely supported languages to include
several regional and non-scheduled Indian languages.
iv.
The model is freely
and publicly available on Hugging Face under an MIT licence.
2. Purpose & Key Advantage:
i.
SraVaani seeks to address
the underrepresentation of regional and non-scheduled Indian languages
in existing speech technologies.
ii.
More than 40 of the
languages covered by the model are not officially supported by existing
speech-recognition systems.
iii.
Its multilingual approach
can help extend speech-based AI technologies to a wider range of India's
linguistic communities.
3. Significance:
i.
Linguistic
Inclusion: Expands speech-recognition
capabilities beyond scheduled and widely supported Indian languages.
ii.
Inclusive AI: Helps bring speech-based artificial intelligence to
languages and dialects that remain underserved by existing technologies.
iii.
Open Accessibility: Its availability under an MIT licence enables
free public access to the model.
iv.
Indian-language AI: Demonstrates efforts to develop AI systems capable of
handling India's extensive linguistic diversity.
Source: THE HINDU - https://www.thehindu.com/news/national/karnataka/first-multilingual-indian-speech-recognition-model-trained-on-65-languages-and-dialects-released/article71341129.ece