DewaterNet: A fusion adversarial real underwater image enhancement network
作者：
Highlights：
• We propose the simple and effective fusion adversarial network for enhancing real underwater image.
• We employ the multi-term objective function to correct color casts.
• We provide visually promising results by the numerous experiments.
• We conduct the ablation study to show the effect of each component and loss component.
摘要
•We propose the simple and effective fusion adversarial network for enhancing real underwater image.•We employ the multi-term objective function to correct color casts.•We provide visually promising results by the numerous experiments.•We conduct the ablation study to show the effect of each component and loss component.
论文关键词：Real underwater image enhancement,Generative adversarial network,Benchmark dataset,Deep learning
论文评审过程：Received 29 July 2020, Revised 22 March 2021, Accepted 25 March 2021, Available online 16 April 2021, Version of Record 18 April 2021.
论文官网地址：https://doi.org/10.1016/j.image.2021.116248
