{"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/shift-net-image-inpainting-via-deep-feature","title":"Shift-Net: Image Inpainting via Deep Feature Rearrangement","arxiv_id":"1801.09392","date":"2018-01-29","proceeding":"ECCV 2018 9","authors":["Zhaoyi Yan","Xiaoming Li","Mu Li","WangMeng Zuo","Shiguang Shan"],"abstract":"Deep convolutional networks (CNNs) have exhibited their potential in image\ninpainting for producing plausible results. However, in most existing methods,\ne.g., context encoder, the missing parts are predicted by propagating the\nsurrounding convolutional features through a fully connected layer, which\nintends to produce semantically plausible but blurry result. In this paper, we\nintroduce a special shift-connection layer to the U-Net architecture, namely\nShift-Net, for filling in missing regions of any shape with sharp structures\nand fine-detailed textures. To this end, the encoder feature of the known\nregion is shifted to serve as an estimation of the missing parts. A guidance\nloss is introduced on decoder feature to minimize the distance between the\ndecoder feature after fully connected layer and the ground-truth encoder\nfeature of the missing parts. With such constraint, the decoder feature in\nmissing region can be used to guide the shift of encoder feature in known\nregion. An end-to-end learning algorithm is further developed to train the\nShift-Net. Experiments on the Paris StreetView and Places datasets demonstrate\nthe efficiency and effectiveness of our Shift-Net in producing sharper,\nfine-detailed, and visually plausible results. The codes and pre-trained models\nare available at https://github.com/Zhaoyi-Yan/Shift-Net.","url_abs":"http://arxiv.org/abs/1801.09392v2","url_pdf":"http://arxiv.org/pdf/1801.09392v2.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":"shift-net-image-inpainting-via-deep-feature","repo_url":"https://github.com/Zhaoyi-Yan/Shift-Net_pytorch","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"shift-net-image-inpainting-via-deep-feature","repo_url":"https://github.com/Zhaoyi-Yan/Shift-Net","is_official":0,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"image-inpainting","task_name":"Image Inpainting"}],"methods":[{"method_slug":"concatenated-skip-connection","method_name":"Concatenated Skip Connection"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"max-pooling","method_name":"Max Pooling"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"u-net","method_name":"U-Net"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1801.09392","atlas_url":"https://app.syntology.ai/?focus=1801.09392","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}