Training stereo Magnification

[danperazzo]

Hello, thank you for the wonderfull work. I was trying to train the stereo magnification algorithm on my own using the train script. However, to do this it needs an image directory. I assumed that in your dataset(RealEstate10K) there would be the images, however, it only has the camera poses. How can I obtain the images so that I can try my model so that it reproduces the conditions of your paper?