{"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/coconet-a-deep-neural-network-for-mapping","title":"CocoNet: A deep neural network for mapping pixel coordinates to color values","arxiv_id":"1805.11357","date":"2018-05-29","proceeding":null,"authors":["Paul Andrei Bricman","Radu Tudor Ionescu"],"abstract":"In this paper, we propose a deep neural network approach for mapping the 2D\npixel coordinates in an image to the corresponding Red-Green-Blue (RGB) color\nvalues. The neural network is termed CocoNet, i.e. coordinates-to-color\nnetwork. During the training process, the neural network learns to encode the\ninput image within its layers. More specifically, the network learns a\ncontinuous function that approximates the discrete RGB values sampled over the\ndiscrete 2D pixel locations. At test time, given a 2D pixel coordinate, the\nneural network will output the approximate RGB values of the corresponding\npixel. By considering every 2D pixel location, the network can actually\nreconstruct the entire learned image. It is important to note that we have to\ntrain an individual neural network for each input image, i.e. one network\nencodes a single image only. To the best of our knowledge, we are the first to\npropose a neural approach for encoding images individually, by learning a\nmapping from the 2D pixel coordinate space to the RGB color space. Our neural\nimage encoding approach has various low-level image processing applications\nranging from image encoding, image compression and image denoising to image\nresampling and image completion. We conduct experiments that include both\nquantitative and qualitative results, demonstrating the utility of our approach\nand its superiority over standard baselines, e.g. bilateral filtering or\nbicubic interpolation. Our code is available at\nhttps://github.com/paubric/python-fuse-coconet.","url_abs":"http://arxiv.org/abs/1805.11357v3","url_pdf":"http://arxiv.org/pdf/1805.11357v3.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":"coconet-a-deep-neural-network-for-mapping","repo_url":"https://github.com/paubric/python-fuse-coconet","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"denoising","task_name":"Denoising"},{"task_slug":"image-compression","task_name":"Image Compression"},{"task_slug":"image-denoising","task_name":"Image Denoising"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}