Thanks, Kubra, for sharing your thoughts. I never used Deep image prior. It is a cool idea, yet, its training cost is a barrier. As you mentioned, the total time spent on training is optimized; that's correct. Yet, this training time is incurred for every image, i.e., every image is a training sample, there is no inference.
Furthermore, I had a colleague who used deep prior for RGB-D images . The training cost becomes more severe as the number of dimensions increases.
So, Yes, for 2D images the cost might be manageable, but I would proceed with caution for higher dimensions (e.g., videos). Thanks again for sharing your input :)
 Depth Completion Using a View-constrained Deep Prior