I wanted to make sure I understand the algorithm as this is different a bit from typical DNNs. Assume I have two images with similar in-painting on both and I only have prior for one of them, can I train the DNN to remove the in-painting on one and then used the trained network to remove similar in-painting from other images that I don't have a prior to?
Hi,
I wanted to make sure I understand the algorithm as this is different a bit from typical DNNs. Assume I have two images with similar in-painting on both and I only have prior for one of them, can I train the DNN to remove the in-painting on one and then used the trained network to remove similar in-painting from other images that I don't have a prior to?
Thanks!