English

A Bayesian Perspective on the Deep Image Prior

Computer Vision and Pattern Recognition 2019-04-17 v1 Machine Learning Machine Learning

Abstract

The deep image prior was recently introduced as a prior for natural images. It represents images as the output of a convolutional network with random inputs. For "inference", gradient descent is performed to adjust network parameters to make the output match observations. This approach yields good performance on a range of image reconstruction tasks. We show that the deep image prior is asymptotically equivalent to a stationary Gaussian process prior in the limit as the number of channels in each layer of the network goes to infinity, and derive the corresponding kernel. This informs a Bayesian approach to inference. We show that by conducting posterior inference using stochastic gradient Langevin we avoid the need for early stopping, which is a drawback of the current approach, and improve results for denoising and impainting tasks. We illustrate these intuitions on a number of 1D and 2D signal reconstruction tasks.

Keywords

Cite

@article{arxiv.1904.07457,
  title  = {A Bayesian Perspective on the Deep Image Prior},
  author = {Zezhou Cheng and Matheus Gadelha and Subhransu Maji and Daniel Sheldon},
  journal= {arXiv preprint arXiv:1904.07457},
  year   = {2019}
}

Comments

CVPR 2019

R2 v1 2026-06-23T08:40:49.796Z