{"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/plop-learning-without-forgetting-for","title":"PLOP: Learning without Forgetting for Continual Semantic Segmentation","arxiv_id":"2011.11390","date":"2020-11-23","proceeding":"CVPR 2021 1","authors":["Arthur Douillard","Yifu Chen","Arnaud Dapogny","Matthieu Cord"],"abstract":"Deep learning approaches are nowadays ubiquitously used to tackle computer vision tasks such as semantic segmentation, requiring large datasets and substantial computational power. Continual learning for semantic segmentation (CSS) is an emerging trend that consists in updating an old model by sequentially adding new classes. However, continual learning methods are usually prone to catastrophic forgetting. This issue is further aggravated in CSS where, at each step, old classes from previous iterations are collapsed into the background. In this paper, we propose Local POD, a multi-scale pooling distillation scheme that preserves long- and short-range spatial relationships at feature level. Furthermore, we design an entropy-based pseudo-labelling of the background w.r.t. classes predicted by the old model to deal with background shift and avoid catastrophic forgetting of the old classes. Our approach, called PLOP, significantly outperforms state-of-the-art methods in existing CSS scenarios, as well as in newly proposed challenging benchmarks.","url_abs":"https://arxiv.org/abs/2011.11390v3","url_pdf":"https://arxiv.org/pdf/2011.11390v3.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":"plop-learning-without-forgetting-for","repo_url":"https://github.com/arthurdouillard/CVPR2021_PLOP","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"plop-learning-without-forgetting-for","repo_url":"https://github.com/mostafaelaraby/bacs-continual-semantic-segmentation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"class-incremental-learning","task_name":"Class Incremental Learning"},{"task_slug":"continual-learning","task_name":"Continual Learning"},{"task_slug":"continual-semantic-segmentation","task_name":"Continual Semantic Segmentation"},{"task_slug":"disjoint-10-1","task_name":"Disjoint 10-1"},{"task_slug":"disjoint-15-1","task_name":"Disjoint 15-1"},{"task_slug":"disjoint-15-5","task_name":"Disjoint 15-5"},{"task_slug":"domain-1-1","task_name":"Domain 1-1"},{"task_slug":"domain-11-1","task_name":"Domain 11-1"},{"task_slug":"domain-11-5","task_name":"Domain 11-5"},{"task_slug":"overlapped-10-1","task_name":"Overlapped 10-1"},{"task_slug":"overlapped-100-10","task_name":"Overlapped 100-10"},{"task_slug":"overlapped-100-5","task_name":"Overlapped 100-5"},{"task_slug":"overlapped-100-50","task_name":"Overlapped 100-50"},{"task_slug":"overlapped-15-1","task_name":"Overlapped 15-1"},{"task_slug":"overlapped-15-5","task_name":"Overlapped 15-5"},{"task_slug":null,"task_name":"Overlapped 19-1"},{"task_slug":"overlapped-50-50","task_name":"Overlapped 50-50"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/disjoint-10-1-on-pascal-voc-2012","task":"Disjoint 10-1","dataset":"PASCAL VOC 2012","model":"PLOP","rank_in_archive_order":5,"of":8,"metrics":{"mIoU":"8.4"},"uses_additional_data":false},{"leaderboard":"/sota/disjoint-15-1-on-pascal-voc-2012","task":"Disjoint 15-1","dataset":"PASCAL VOC 2012","model":"PLOP","rank_in_archive_order":6,"of":9,"metrics":{"mIoU":"46.5"},"uses_additional_data":false},{"leaderboard":"/sota/disjoint-15-5-on-pascal-voc-2012","task":"Disjoint 15-5","dataset":"PASCAL VOC 2012","model":"PLOP","rank_in_archive_order":7,"of":9,"metrics":{"Mean IoU":"64.3"},"uses_additional_data":false},{"leaderboard":"/sota/overlapped-10-1-on-pascal-voc-2012","task":"Overlapped 10-1","dataset":"PASCAL VOC 2012","model":"PLOP","rank_in_archive_order":9,"of":13,"metrics":{"mIoU":"30.45"},"uses_additional_data":false},{"leaderboard":"/sota/overlapped-100-10-on-ade20k","task":"Overlapped 100-10","dataset":"ADE20K","model":"PLOP","rank_in_archive_order":5,"of":6,"metrics":{"Mean IoU (test) ":"31.59"},"uses_additional_data":false},{"leaderboard":"/sota/overlapped-100-10-on-ade20k","task":"Overlapped 100-10","dataset":"ADE20K","model":"MiB","rank_in_archive_order":6,"of":6,"metrics":{"Mean IoU (test) ":"29.24"},"uses_additional_data":false},{"leaderboard":"/sota/overlapped-100-5-on-ade20k","task":"Overlapped 100-5","dataset":"ADE20K","model":"PLOP","rank_in_archive_order":6,"of":8,"metrics":{"mIoU":"28.75"},"uses_additional_data":false},{"leaderboard":"/sota/overlapped-100-5-on-ade20k","task":"Overlapped 100-5","dataset":"ADE20K","model":"MiB","rank_in_archive_order":7,"of":8,"metrics":{"mIoU":"25.96"},"uses_additional_data":false},{"leaderboard":"/sota/overlapped-100-50-on-ade20k","task":"Overlapped 100-50","dataset":"ADE20K","model":"PLOP","rank_in_archive_order":6,"of":7,"metrics":{"mIoU":"32.94"},"uses_additional_data":false},{"leaderboard":"/sota/overlapped-100-50-on-ade20k","task":"Overlapped 100-50","dataset":"ADE20K","model":"MiB","rank_in_archive_order":7,"of":7,"metrics":{"mIoU":"32.79"},"uses_additional_data":false},{"leaderboard":"/sota/overlapped-15-1-on-pascal-voc-2012","task":"Overlapped 15-1","dataset":"PASCAL VOC 2012","model":"PLOP","rank_in_archive_order":9,"of":13,"metrics":{"mIoU":"54.64"},"uses_additional_data":false},{"leaderboard":"/sota/overlapped-15-1-on-pascal-voc-2012","task":"Overlapped 15-1","dataset":"PASCAL VOC 2012","model":"MiB","rank_in_archive_order":11,"of":13,"metrics":{"mIoU":"29.29"},"uses_additional_data":false},{"leaderboard":"/sota/overlapped-15-5-on-pascal-voc-2012","task":"Overlapped 15-5","dataset":"PASCAL VOC 2012","model":"PLOP","rank_in_archive_order":9,"of":13,"metrics":{"Mean IoU (val)":"70.09"},"uses_additional_data":false},{"leaderboard":"/sota/overlapped-15-5-on-pascal-voc-2012","task":"Overlapped 15-5","dataset":"PASCAL VOC 2012","model":"MiB","rank_in_archive_order":10,"of":13,"metrics":{"Mean IoU (val)":"70.08"},"uses_additional_data":false},{"leaderboard":"/sota/overlapped-50-50-on-ade20k","task":"Overlapped 50-50","dataset":"ADE20K","model":"PLOP","rank_in_archive_order":4,"of":7,"metrics":{"mIoU":"30.4"},"uses_additional_data":false},{"leaderboard":"/sota/overlapped-50-50-on-ade20k","task":"Overlapped 50-50","dataset":"ADE20K","model":"MiB","rank_in_archive_order":7,"of":7,"metrics":{"mIoU":"29.31"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2011.11390","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}