Hi, I have some data pairs of low resolution and high resolution, I want to use this ipadapter method to do the image SR task. When training, the low resolution image is encoded, and use the high resolution image as GT. Do you think this is possible? and what is the minimum amount of data required?
Hi, I have some data pairs of low resolution and high resolution, I want to use this ipadapter method to do the image SR task. When training, the low resolution image is encoded, and use the high resolution image as GT. Do you think this is possible? and what is the minimum amount of data required?