{"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/distance-to-center-of-mass-encoding-for","title":"Distance to Center of Mass Encoding for Instance Segmentation","arxiv_id":"1711.09060","date":"2017-11-24","proceeding":null,"authors":["Thomio Watanabe","Denis Wolf"],"abstract":"The instance segmentation can be considered an extension of the object\ndetection problem where bounding boxes are replaced by object contours.\nStrictly speaking the problem requires to identify each pixel instance and\nclass independently of the artifice used for this mean. The advantage of\ninstance segmentation over the usual object detection lies in the precise\ndelineation of objects improving object localization. Additionally, object\ncontours allow the evaluation of partial occlusion with basic image processing\nalgorithms. This work approaches the instance segmentation problem as an\nannotation problem and presents a novel technique to encode and decode ground\ntruth annotations. We propose a mathematical representation of instances that\nany deep semantic segmentation model can learn and generalize. Each individual\ninstance is represented by a center of mass and a field of vectors pointing to\nit. This encoding technique has been denominated Distance to Center of Mass\nEncoding (DCME).","url_abs":"http://arxiv.org/abs/1711.09060v1","url_pdf":"http://arxiv.org/pdf/1711.09060v1.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":"distance-to-center-of-mass-encoding-for","repo_url":"https://github.com/usp-lrm/dcme","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"caffe2","reach":null}],"tasks":[{"task_slug":"instance-segmentation","task_name":"Instance Segmentation"},{"task_slug":"object","task_name":"Object"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"object-localization","task_name":"Object Localization"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"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}