{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/maanet-multi-view-aware-attention-networks","title":"MAANet: Multi-view Aware Attention Networks for Image Super-Resolution","arxiv_id":"1904.06252","date":"2019-04-12","proceeding":null,"authors":["Jingcai Guo","Shiheng Ma","Song Guo"],"abstract":"In most recent years, deep convolutional neural networks (DCNNs) based image\nsuper-resolution (SR) has gained increasing attention in multimedia and\ncomputer vision communities, focusing on restoring the high-resolution (HR)\nimage from a low-resolution (LR) image. However, one nonnegligible flaw of\nDCNNs based methods is that most of them are not able to restore\nhigh-resolution images containing sufficient high-frequency information from\nlow-resolution images with low-frequency information redundancy. Worse still,\nas the depth of DCNNs increases, the training easily encounters the problem of\nvanishing gradients, which makes the training more difficult. These problems\nhinder the effectiveness of DCNNs in image SR task. To solve these problems, we\npropose the Multi-view Aware Attention Networks (MAANet) for image SR task.\nSpecifically, we propose the local aware (LA) and global aware (GA) attention\nto deal with LR features in unequal manners, which can highlight the\nhigh-frequency components and discriminate each feature from LR images in the\nlocal and the global views, respectively. Furthermore, we propose the local\nattentive residual-dense (LARD) block, which combines the LA attention with\nmultiple residual and dense connections, to fit a deeper yet easy to train\narchitecture. The experimental results show that our proposed approach can\nachieve remarkable performance compared with other state-of-the-art methods.","url_abs":"http://arxiv.org/abs/1904.06252v1","url_pdf":"http://arxiv.org/pdf/1904.06252v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"maanet-multi-view-aware-attention-networks","repo_url":"https://github.com/AdivarekarBhumit/MAANet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"image-super-resolution","task_name":"Image Super-Resolution"},{"task_slug":"super-resolution","task_name":"Super-Resolution"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}