search

acvictor / Obama-Lip-Sync

An implementation of ObamaNet: Photo-realistic lip-sync from text.
MIT License
124 stars 33 forks source link

Mapping dlib mouth points to image for pix2pix model #17

Closed waffel183 closed 3 years ago

waffel183 commented 3 years ago

Without the premade pix2pix model originally used in this repo, I have to make my own model. How can I put the points given in the map "rawData" to the images of Obama with the mouth cut out? similar to the frames of "results/input500ms.mp4"

Alternatively, could you upload your own pix2pix model?

waffel183 commented 3 years ago

It is also possible I'm going about it the wrong way, if so, could you point me in the right direction?

waffel183 commented 3 years ago

Found the solution, create an "output" folder in de main folder. the script was supposed to do this by itself, but maybe because I was using a venv it didn't work.