A GAN-based input-size flexibility model for single image dehazing

作者：

Highlights：

• First, an end-to-end input-size flexibility cGAN model is proposed for single image dehazing.

• Second, a simple and effective UR-Net structure is designed based on the popular U-Net structure and residual learning.

• Third, a consistency loss is proposed to keep the transformation consistency between dehazing image and real image.

摘要

•First, an end-to-end input-size flexibility cGAN model is proposed for single image dehazing.•Second, a simple and effective UR-Net structure is designed based on the popular U-Net structure and residual learning.•Third, a consistency loss is proposed to keep the transformation consistency between dehazing image and real image.

论文关键词：Generative adversarial network,Image dehazing,Image restoration

论文评审过程：Received 11 June 2021, Revised 29 November 2021, Accepted 30 November 2021, Available online 14 December 2021, Version of Record 27 December 2021.

论文官网地址：https://doi.org/10.1016/j.image.2021.116599