Bocca
A push-to-talk tool that transforms your audio into text
94 followers
Bocca transforms your voice into perfectly transcribed text in seconds, revolutionizing your workflow with an intuitive push-to-talk experience. Handled by AI.
Subscribe
Sign in
Bocca
@paulmarniquet Interesting idea, and the best part is it's offline. quick question how well it handle messy speech or background noise? I tend to ramble 😅
Bocca
@hamza_afzal_butt Hey! Thanks a lot :) As for all that rambling, the AI models handle this relatively well. Background noise can sometimes be a problem in terms of comprehension, but it's still pretty well managed, and if it isn't, you can set up another audio device to have less of it.
@paulmarniquet Thanks :)
@paulmarniquet How does Bocca handle specialized vocabularies (like medical or legal terms)? Can users train or fine-tune models for niche use cases?
Bocca
@masump Hello !
The current Whisper models from OpenAI used by Bocca already offer excellent recognition of common technical terms.
In upcoming versions of Bocca, we plan to implement a specialized dictionary feature allowing each user to add specific vocabulary words that the models might have difficulty recognizing.
Instant transcripts save so much time! 😄
That’s such a great idea — I also really love the design and overall concept. By the way, does it handle dialects well too? Haha 😄
Bocca
@kay_arkain Thanks for your kind words! 😊
About dialects, Bocca handles most accents pretty well thanks to Whisper's training. Some quirky dialects might trip it up occasionally, but we're working on that ! The custom dictionary feature we're adding should help smooth things out too.