search

andrerochow / fsrt

Official implementation of the CVPR 2024 paper "FSRT: Facial Scene Representation Transformer for Face Reenactment from Factorized Appearance, Head-pose, and Facial Expression Features"
https://andrerochow.github.io/fsrt
67 stars 4 forks source link

Output contains both Driving and Output video as single output file #8

Open nitinmukesh opened 3 weeks ago

nitinmukesh commented 3 weeks ago

How can we output only the output video without combining Driving video

johndpope commented 3 weeks ago

how bout make the driving video also the input video?

Inferencer commented 3 weeks ago

How can we output only the output video without combining Driving video

I cant remember what I did but i fixed that in this repo probs the demo.py concat stuff so you could copy that code

nitinmukesh commented 2 weeks ago

How can we output only the output video without combining Driving video

I cant remember what I did but i fixed that in this repo probs the demo.py concat stuff so you could copy that code

Sorry which repo should I refer

This is the code in demo.py

    imageio.mimsave(opt.result_video, [img_as_ubyte(frame) for frame in predictions], fps=20)
#    imageio.mimsave(opt.result_video, [img_as_ubyte(np.concatenate([source_image[0],frame], axis=1)) for frame in predictions], fps=20)
Inferencer commented 2 weeks ago

How can we output only the output video without combining Driving video

I cant remember what I did but i fixed that in this repo probs the demo.py concat stuff so you could copy that code

Sorry which repo should I refer

This is the code in demo.py

    imageio.mimsave(opt.result_video, [img_as_ubyte(frame) for frame in predictions], fps=20)
#    imageio.mimsave(opt.result_video, [img_as_ubyte(np.concatenate([source_image[0],frame], axis=1)) for frame in predictions], fps=20)

my bad forgot to add link https://github.com/Inferencer/SickFace i think i removed the axis stuff tho