Computational imaging is the eye of almost all intelligent visual systems, which optically encodes 3D scene information and computationally decodes it for scene reconstruction and visual perception. Recently computational imaging methods, together with deep neural networks (and in particular generative models) have been successfully applied for image reconstruction and enhancement for mobile phones. In this talk, I will briefly go over some of our recent work in which computational imaging with generative priors being employed for large-scale image super-resolution, image denoising, removing lens flares for under-display cameras, depth sensing, and lensless imaging. I hope some of these methods can also be applied for medical imaging in near future with the collaboration in the TRS project.