NAS-DIP: Learning Deep Image Prior with Neural Architecture Search

In European Conference on Computer Vision (ECCV), 2020

Recent work has shown that the structure of deep convolutional neural networks can be used as a structured image prior for solving various inverse image restoration tasks. Instead of using hand-designed architectures, we propose to search for neural architectures that capture stronger image priors. Building upon a generic U-Net architecture, our core contribution lies in designing new search spaces for (1) an upsampling cell and (2) a pattern of cross-scale residual connections. We search for an improved network by leveraging an existing neural architecture search algorithm (using reinforcement learning with a recurrent neural network controller). We validate the effectiveness of our method via a wide variety of applications, including image restoration, dehazing, image-to-image translation, and matrix factorization. Extensive experimental results show that our algorithm performs favorably against state-of-the-art learning-free approaches and reaches competitive performance with existing learning-based methods in some cases.


ECCV 2020

Yun-Chun Chen*, Chen Gao*, Esther Robb, and Jia-Bin Huang, "NAS-DIP: Learning Deep Image Prior with Neural Architecture Search", in European Conference on Computer Vision, 2020.

*: equal contribution

    author    = {Chen, Yun-Chun and Gao, Chen and Robb, Esther and Huang, Jia-Bin},
    title     = {NAS-DIP: Learning Deep Image Prior with Neural Architecture Search},
    booktitle = {European Conference on Computer Vision (ECCV)},
    year      = {2020}

Network Architecture

Overview of NAS-DIP