Our | |
approach matches the representation of an image view containing randomly masked patches to the representation of the original | |
unmasked image. |
Our | |
approach matches the representation of an image view containing randomly masked patches to the representation of the original | |
unmasked image. |