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.
@article{arxiv.2008.11713,
title = {NAS-DIP: Learning Deep Image Prior with Neural Architecture Search},
author = {Yun-Chun Chen and Chen Gao and Esther Robb and Jia-Bin Huang},
journal= {arXiv preprint arXiv:2008.11713},
year = {2020}
}
Comments
ECCV 2020. Project: https://yunchunchen.github.io/NAS-DIP/ Code: https://github.com/YunChunChen/NAS-DIP-pytorch The first two authors contributed equally to this work