UnfairGAN: An Enhanced Generative Adversarial Network for Raindrop Removal from A Single Image
This addresses a domain-specific challenge for autonomous vehicles and machine learning systems in bad weather, but it is incremental as it builds on existing GAN approaches.
The paper tackles the problem of removing raindrops from single images, which can impair vision in applications like autonomous vehicles, by proposing UnfairGAN, an enhanced generative adversarial network that uses prior high-level information like edges and rain estimation. The results show it outperforms other state-of-the-art methods in quantitative metrics and visual quality.
Image deraining is a new challenging problem in real-world applications, such as autonomous vehicles. In a bad weather condition of heavy rainfall, raindrops, mainly hitting glasses or windshields, can significantly reduce observation ability. Moreover, raindrops spreading over the glass can yield refraction's physical effect, which seriously impedes the sightline or undermine machine learning systems. In this paper, we propose an enhanced generative adversarial network to deal with the challenging problems of raindrops. UnfairGAN is an enhanced generative adversarial network that can utilize prior high-level information, such as edges and rain estimation, to boost deraining performance. To demonstrate UnfairGAN, we introduce a large dataset for training deep learning models of rain removal. The experimental results show that our proposed method is superior to other state-of-the-art approaches of deraining raindrops regarding quantitative metrics and visual quality.