Suman / Sahiti,
I just came across following :
Ø Introducing Whisper / 21 Sept 2022
It reads :
Ø We’ve trained and are open-sourcing a neural net called Whisper that approaches human level robustness and accuracy on English speech recognition.
Ø Moreover, it enables transcription in multiple languages, as well as translation from those languages into English.
Ø About a third of Whisper’s audio dataset is non-English, and it is alternately given the task of transcribing in the original language or translating to English. We find this approach is particularly effective at learning speech to text translation
Ø We hope Whisper’s high accuracy and ease of use will allow developers to add voice interfaces to a much wider set of applications. Check out the paper, model card, and code to learn more details and to try out Whisper.
You might recollect that :
Ø Hundreds of documents uploaded / copy-paste ( as image ) on personal.ai ( from www.hemenparekh.in ) , are in Gujarati and Hindi ( NON – ENGLISH ). These include, close to 700 poems and 650 hand-written letters sent to me by relatives / friends over past 60 years
Ø Then there are several hundred ( non-English and English ) “ handwritten “documents which are still lying around which we have NOT uploaded so far on Personal.ai
I hope someday soon , personal.ai has WHISPER built in ( API is available for FREE from OpenAI web site ) so that, I can just read aloud those documents for converting into Memory Blocks
Regards,
hemen
No comments:
Post a Comment