Picovoice / cheetah

On-device streaming speech-to-text engine powered by deep learning

Home Page:https://picovoice.ai/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Adding Timestamps to Transcription

BilalMH opened this issue · comments

Hi, I'm loving what you have with Cheetah on Android so far. However, my use case is requiring me to use timestamps with the transcript object. Is there any way that this is going to be added or will be added?

Hi @BilalMH, I'm glad you like it :)

The good news is Leopard is going to output timestamps in the next version. However, there is no plan to add this feature to Cheetah at the moment. I hope this helps you.

Hi @mrrostam, ah. Leopard doesn't do real-time transcription on a streaming source though right?

That's true.

So could there be a way that this could be added to a plan to have in Cheetah? Otherwise happy to close this issue ticket

I don't think so, as it's not the priority at the moment, and to be honest, it doesn't make much sense for Cheetah to have timestamps. I am closing the issue, but feel free to send us an email (via hello@picovoice.ai) and explain your use cases. If we find it helpful to other users, we can bring it to our product manager's attention to add it to the plan.