×

Bhasa

Speech to Text AI that auto tunes to the voices in the conversation

Bhasa

Speech to Text AI that auto tunes to the voices in the conversation

Bhasa aspires to provide seamless multilingual Speech Recognition AI that can tune to the individual voices in any conversation. This provides for a substantial bump in speech to text conversion accuracy. We expect the product to be of much help to transcribe meeting conversations, create Youtube captions and to dictate notes.

The Beta version offers speech recognition for English audio. Better language coverage and multilingual support post beta. We have deployed couple of models for the beta launch, they easily outperform the giants from Redmond and Mountainview. But the real deal is when Bhasa tunes to your voice. Give us a shout on Twitter, we will have Bhasa tuned for you!

looking good but taking time
@ankitsharmabpl hi Ankit, thanks for trying out the tool. Mind telling us where you find the product slow.
I tried this product on a YT video. (https://www.youtube.com/watch?v=_lU91279xZk&t=435s)

Pro:
- It worked
- It was fast (30 seconds is completely fine)
- No downloads or hacks into my computer to make it work.
- Pricing is reasonable.

Cons:
- It is one, giant, run-on set of all lower-case text. While they have the auto on top so you can follow along, still.
- A lot of spelling and space errors, however, this kind of problem is prevalent in all of these efforts, more or less. When compared to the big boys, as it is called, they do a reasonable job.
- not usable, yet, for meeting transcription as it misses some people talking in the same YT video. This is not the end of the world, but you should know this.

Overall, this is a damn good effort that the developer has put a lot of work into and should be recognized for that effort. This would be a great product to follow if transcription is your thing. For a beta? Outstanding first effort. Well-done.

Join to leave a comment