{"id":16482,"date":"2024-08-09T12:15:50","date_gmt":"2024-08-09T10:15:50","guid":{"rendered":"https:\/\/www.sling.si\/?post_type=wpdmpro&p=16482"},"modified":"2025-07-16T14:48:45","modified_gmt":"2025-07-16T12:48:45","slug":"super-resolution","status":"publish","type":"wpdmpro","link":"https:\/\/www.sling.si\/en\/download\/super-resolution\/","title":{"rendered":"Super-resolution"},"content":{"rendered":"

Many breakthroughs in speed and accuracy of single image super-resolution (SISR) have been achieved. One of the biggest challenges is how to recover finer texture details when super-resolution is applied at large upscaling factors. A typical solution to SISR involves using a convolutional neural network (CNN), however new approaches using a generative adversarial network (GAN) are now also popular. The behavior of optimization-based super-resolution methods is principally driven by the choice of the objective function. In this work, we present an evaluation of SRResNet and SRGAN. SRResNet is a deep residual network and SRGAN is a generative adversarial network for image super-resolution (SR). SRResNet is able to recover reasonable quality photo-realistic textures from heavily downsampled images. SRGAN is capable of inferring photo-realistic natural images with high upscaling factors. This is achieved using a perceptual loss function which consists of an adversarial loss and a content loss. The adversarial loss pushes the solution to the natural image manifold using a discriminator network that is trained to differentiate between the super-resolved images and original photo-realistic images. The content loss is motivated by perceptual similarity instead of similarity in pixel space.<\/p>\n

Download the entire article:<\/p>\n

\"Download

<\/div><\/a><\/p>\n

Author: Sebastien Strban; University of Ljubljana<\/p>\n","protected":false},"excerpt":{"rendered":"

Many breakthroughs in speed and accuracy of single image super-resolution (SISR) have been achieved. One of the biggest challenges is how to recover finer texture details when super-resolution is applied …<\/p>\n","protected":false},"author":13,"featured_media":16356,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_acf_changed":false,"__wpdm_changelog":[]},"wpdmcategory":[635],"wpdmtag":[485,633,468,632,634,636,637,657],"class_list":["post-16482","wpdmpro","type-wpdmpro","status-publish","has-post-thumbnail","hentry","wpdmcategory-scientific-article","wpdmtag-fri-ul","wpdmtag-generative-adversarial-network","wpdmtag-hpc","wpdmtag-sisr","wpdmtag-srgan","wpdmtag-super-resolution","wpdmtag-supercomputing","wpdmtag-ul"],"acf":[],"aioseo_notices":[],"aioseo_head":"\n\t\t\n\t\n\t\n\t\n\t\n\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t