Midline cut issue with multi-voice

ArturoAmaya / ScriptParser

A *VERY* basic python script parser for HeyGen scripts

1 stars 0 forks source link

Midline cut issue with multi-voice #1

Open ArturoAmaya opened 5 months ago

ArturoAmaya commented 5 months ago

Issue where the midline cut points are not properly calculated for midline cuts that transition between multiple voices

ArturoAmaya commented 5 months ago

Ok it seems that I can't just get this right with the provided captions, meaning I would have to load in the whisper model from openAI and hope beyond hope that I can get it to give me really good timestamps. I'm not sure what to do about it.

ArturoAmaya commented 5 months ago

For now I am simply going to say no midline voice transitions are allowed so we can keep the program lightweight and not use an AI model which will slow things down but not give us better performance.