{"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/empty-cities-image-inpainting-for-a-dynamic","title":"Empty Cities: Image Inpainting for a Dynamic-Object-Invariant Space","arxiv_id":"1809.10239","date":"2018-09-20","proceeding":null,"authors":["Berta Bescos","José Neira","Roland Siegwart","Cesar Cadena"],"abstract":"In this paper we present an end-to-end deep learning framework to turn images\nthat show dynamic content, such as vehicles or pedestrians, into realistic\nstatic frames. This objective encounters two main challenges: detecting all the\ndynamic objects, and inpainting the static occluded background with plausible\nimagery. The second problem is approached with a conditional generative\nadversarial model that, taking as input the original dynamic image and its\ndynamic/static binary mask, is capable of generating the final static image.\nThe former challenge is addressed by the use of a convolutional network that\nlearns a multi-class semantic segmentation of the image.\n  These generated images can be used for applications such as augmented reality\nor vision-based robot localization purposes. To validate our approach, we show\nboth qualitative and quantitative comparisons against other state-of-the-art\ninpainting methods by removing the dynamic objects and hallucinating the static\nstructure behind them. Furthermore, to demonstrate the potential of our\nresults, we carry out pilot experiments that show the benefits of our proposal\nfor visual place recognition.","url_abs":"http://arxiv.org/abs/1809.10239v2","url_pdf":"http://arxiv.org/pdf/1809.10239v2.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":"empty-cities-image-inpainting-for-a-dynamic","repo_url":"https://github.com/bertabescos/EmptyCities","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"image-inpainting","task_name":"Image Inpainting"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"},{"task_slug":"visual-place-recognition","task_name":"Visual Place Recognition"}],"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}