Open j-sieger opened 4 years ago
once you have a trained SinGAN model, it is subtitle for generation of the image you trained it for. For editing and harmonization you can input different reference images (as described in the readme)
@tamarott Does this mean that we can't used a SR model that is trained on imageA to super-resolve for imageB, C, D...etc.?
Once we train the SR GAN / EditorGAN/Animator/Hormonization , whether we can run this inference on random images from google and get the corresponding output? Example: i have trained animator with tree image and got the nice animation of wind blow. If i use these weights and run the generator with another image say flight or something. whether i can get animation?