You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hey folks! I posted here a while back about a demo of TalkNet which is now linked in the repository README. We took that demo and turned it into a full-fledged functionality that's been hot lately with products like Opus.
Basically, we take in any content and can reformat it to a given aspect ratio based on who's actively speaking, how many people are in the scene, etc. This automates a lot of video creation processes that people take for granted when it comes to publishing social content.
Hey folks! I posted here a while back about a demo of TalkNet which is now linked in the repository README. We took that demo and turned it into a full-fledged functionality that's been hot lately with products like Opus.
Basically, we take in any content and can reformat it to a given aspect ratio based on who's actively speaking, how many people are in the scene, etc. This automates a lot of video creation processes that people take for granted when it comes to publishing social content.
Check it out here: https://www.sievedata.com/functions/sieve/autocrop
Here's an example video which we "reformated": https://www.youtube.com/watch?v=4xACV5qm0Io
c589b43f-cb92-444c-b385-2b5913769038-output-fa1b5949-c9b1-4539-ae5b-0ac0946e1ec9.mp4
Thanks again Tao for your great contributions! We're really happy to be building on top of it :)
The text was updated successfully, but these errors were encountered: