At present, there has been significant research in the field of image super-resolution technology, and various research approaches have been proposed for super-resolution algorithms. Whether it is traditional algorithms, machine learning algorithms, or research approaches based on deep learning convolutional neural networks (CNN) and generative adversarial networks (GAN), remarkable results and achievements have been made. However, there still exists the issue that upsampling images often leads to over-smoothing and a lack of high-frequency details. SRGAN, by introducing adversarial loss and perceptual loss, can effectively restore the texture and details of images. Its generator employs a deep residual network structure, capable of recovering realistic high-resolution images from low-resolution ones, while the discriminator enhances the performance of the generator by distinguishing between real and generated images. Experimental results indicate that SRGAN outperforms traditional methods in terms of PSNR and SSIM metrics on the DIV2K dataset.
Yao et al. (Tue,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: