Closed tcharlss closed 3 years ago
You can try the implementation in my fork of GIMP-ML, which should work as expected: https://github.com/valgur/GIMP-ML-Hub
Thanks I will give it a try :)
@tcharlss MiDas was originally trained by the authors on images of size 384 by 384.
The MonoDepth filter is super handy, but it seems it operates at a very low resolution and then resize the new layer, making it quite blurry, and thus not so usable.
Could it be possible to add a parameter in order to increase its resolution?