Elsaam2y / DINet_optimized

An optimized pipeline for DINet reducing inference latency for up to 60% 🚀. Kudos for the authors of the original repo for this amazing work.
93 stars 15 forks source link

How to get a video.csv file for a custom video in Conda ? #9

Closed Neurophars closed 7 months ago

davidmartinrius commented 9 months ago

You mean you want to generate face landmarks with openface installed in conda? Openface without docker is a headache to install.

I don't know if you mean this because I need more details... What do you need exactly?

Neurophars commented 9 months ago

Вы имеете в виду, что хотите создавать ориентиры на лицах с помощью openface, установленного в Конде? Openface без докера — головная боль при установке.

Я не знаю, имеете ли вы это в виду, потому что мне нужно больше подробностей... Что именно вам нужно?

Thanks for the quick reply, I already figured out how to make video.csv. Is it possible to make the mask like in my attached picture? Or is the mask limited to a square? mask

davidmartinrius commented 9 months ago

As far as I know this is limited to a square/rectangle. With Dinet implementation cheeks won't move.

Neurophars commented 9 months ago

As far as I know this is limited to a square/rectangle. With Dinet implementation cheeks won't move.

Got it, thanks.

Elsaam2y commented 7 months ago

Feel free to reopen the issue if needed.