GAN-Based Image Super-Resolution with a Novel Quality Loss
GAN-Based Image Super-Resolution with a Novel Quality Loss
复制标题
基于 GAN 的图像超分辨率，具有新颖的质量损失
DOI：
10.1155/2020/5217429
复制
发表时间：
2020-02-18
影响因子：
--
通讯作者：
Zhao, Shengjie
中科院分区：
文献类型：
--
作者：
Zhu, Xining;Zhang, Lin;Zhao, Shengjie
单图像超分辨率（SISR）近年来一直是一个非常有吸引力的研究课题。由于深度学习和生成对抗网络（GANs），SISR已经取得了突破。然而，生成的图像仍然存在不理想的伪影。在本文中，我们针对SISR任务提出了一种名为GMGAN的新方法。在该方法中，为了生成更符合人类视觉系统（HVS）的图像，我们通过整合一种名为梯度幅度相似性偏差（GMSD）的图像质量评估（IQA）指标来设计一种质量损失。据我们所知，这是首次真正将一种IQA指标整合到SISR中。此外，为了克服原始GAN的不稳定性，我们使用了一种GAN的变体，即改进的Wasserstein GANs（WGAN - GP）训练方法。除了GMGAN，我们还强调了训练数据集的重要性。实验表明，具有质量损失和WGAN - GP的GMGAN能够生成视觉上吸引人的结果，并达到了新的先进水平。此外，大量具有丰富纹理的高质量训练图像对结果有益。
Single image super-resolution (SISR) has been a very attractive research topic in recent years. Breakthroughs in SISR have been achieved due to deep learning and generative adversarial networks (GANs). However, the generated image still suffers from undesired artifacts. In this paper, we propose a new method named GMGAN for SISR tasks. In this method, to generate images more in line with human vision system (HVS), we design a quality loss by integrating an image quality assessment (IQA) metric named gradient magnitude similarity deviation (GMSD). To our knowledge, it is the first time to truly integrate an IQA metric into SISR. Moreover, to overcome the instability of the original GAN, we use a variant of GANs named improved training of Wasserstein GANs (WGAN-GP). Besides GMGAN, we highlight the importance of training datasets. Experiments show that GMGAN with quality loss and WGAN-GP can generate visually appealing results and set a new state of the art. In addition, large quantity of high-quality training images with rich textures can benefit the results.