{"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/you-only-look-twice-rapid-multi-scale-object","title":"You Only Look Twice: Rapid Multi-Scale Object Detection In Satellite Imagery","arxiv_id":"1805.09512","date":"2018-05-24","proceeding":null,"authors":["Adam Van Etten"],"abstract":"Detection of small objects in large swaths of imagery is one of the primary\nproblems in satellite imagery analytics. While object detection in ground-based\nimagery has benefited from research into new deep learning approaches,\ntransitioning such technology to overhead imagery is nontrivial. Among the\nchallenges is the sheer number of pixels and geographic extent per image: a\nsingle DigitalGlobe satellite image encompasses >64 km2 and over 250 million\npixels. Another challenge is that objects of interest are minuscule (often only\n~10 pixels in extent), which complicates traditional computer vision\ntechniques. To address these issues, we propose a pipeline (You Only Look\nTwice, or YOLT) that evaluates satellite images of arbitrary size at a rate of\n>0.5 km2/s. The proposed approach can rapidly detect objects of vastly\ndifferent scales with relatively little training data over multiple sensors. We\nevaluate large test images at native resolution, and yield scores of F1 > 0.8\nfor vehicle localization. We further explore resolution and object size\nrequirements by systematically testing the pipeline at decreasing resolution,\nand conclude that objects only ~5 pixels in size can still be localized with\nhigh confidence. Code is available at https://github.com/CosmiQ/yolt.","url_abs":"http://arxiv.org/abs/1805.09512v1","url_pdf":"http://arxiv.org/pdf/1805.09512v1.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":"you-only-look-twice-rapid-multi-scale-object","repo_url":"https://github.com/CosmiQ/yolt","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"you-only-look-twice-rapid-multi-scale-object","repo_url":"https://github.com/avanetten/avanetten.github.io","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}},{"paper_slug":"you-only-look-twice-rapid-multi-scale-object","repo_url":"https://github.com/avanetten/yolt","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"you-only-look-twice-rapid-multi-scale-object","repo_url":"https://github.com/native2019/hello","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"you-only-look-twice-rapid-multi-scale-object","repo_url":"https://github.com/zk2ly/Glass_insulator_defect_detection","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"you-only-look-twice-rapid-multi-scale-object","repo_url":"https://github.com/eavise-kul/lightnet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"you-only-look-twice-rapid-multi-scale-object","repo_url":"https://gitlab.com/eavise/lightnet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1805.09512","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}