{"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/pixelnn-example-based-image-synthesis","title":"PixelNN: Example-based Image Synthesis","arxiv_id":"1708.05349","date":"2017-08-17","proceeding":"ICLR 2018 1","authors":["Aayush Bansal","Yaser Sheikh","Deva Ramanan"],"abstract":"We present a simple nearest-neighbor (NN) approach that synthesizes\nhigh-frequency photorealistic images from an \"incomplete\" signal such as a\nlow-resolution image, a surface normal map, or edges. Current state-of-the-art\ndeep generative models designed for such conditional image synthesis lack two\nimportant things: (1) they are unable to generate a large set of diverse\noutputs, due to the mode collapse problem. (2) they are not interpretable,\nmaking it difficult to control the synthesized output. We demonstrate that NN\napproaches potentially address such limitations, but suffer in accuracy on\nsmall datasets. We design a simple pipeline that combines the best of both\nworlds: the first stage uses a convolutional neural network (CNN) to maps the\ninput to a (overly-smoothed) image, and the second stage uses a pixel-wise\nnearest neighbor method to map the smoothed output to multiple high-quality,\nhigh-frequency outputs in a controllable manner. We demonstrate our approach\nfor various input modalities, and for various domains ranging from human faces\nto cats-and-dogs to shoes and handbags.","url_abs":"http://arxiv.org/abs/1708.05349v1","url_pdf":"http://arxiv.org/pdf/1708.05349v1.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":"pixelnn-example-based-image-synthesis","repo_url":"https://github.com/aayushbansal/PixelNN-Code","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"image-generation","task_name":"Image Generation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1708.05349","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}