Papers › PLOP: Learning without Forgetting for Continual Semantic Segmentation

PLOP: Learning without Forgetting for Continual Semantic Segmentation

23 Nov 2020CVPR 2021 1arXiv:2011.11390archive 2025-07-28

Arthur Douillard, Yifu Chen, Arnaud Dapogny, Matthieu Cord

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.

PaperPDFConference PDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

arthurdouillard/CVPR2021_PLOP officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Class Incremental LearningContinual LearningContinual Semantic SegmentationDisjoint 10-1Disjoint 15-1Disjoint 15-5Domain 1-1Domain 11-1Domain 11-5Overlapped 10-1Overlapped 100-10Overlapped 100-5Overlapped 100-50Overlapped 15-1Overlapped 15-5Overlapped 50-50SegmentationSemantic Segmentation

1 archive task tag without a task page not shown.

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Disjoint 10-1 PASCAL VOC 2012 PLOP mIoU 8.4 #5 of 8 Archive leaderboard report
Disjoint 15-1 PASCAL VOC 2012 PLOP mIoU 46.5 #6 of 9 Archive leaderboard report
Disjoint 15-5 PASCAL VOC 2012 PLOP Mean IoU 64.3 #7 of 9 Archive leaderboard report
Overlapped 10-1 PASCAL VOC 2012 PLOP mIoU 30.45 #9 of 13 Archive leaderboard report
Overlapped 100-10 ADE20K PLOP Mean IoU (test) 31.59 #5 of 6 Archive leaderboard report
Overlapped 100-10 ADE20K MiB Mean IoU (test) 29.24 #6 of 6 Archive leaderboard report
Overlapped 100-5 ADE20K PLOP mIoU 28.75 #6 of 8 Archive leaderboard report
Overlapped 100-5 ADE20K MiB mIoU 25.96 #7 of 8 Archive leaderboard report
Overlapped 100-50 ADE20K PLOP mIoU 32.94 #6 of 7 Archive leaderboard report
Overlapped 100-50 ADE20K MiB mIoU 32.79 #7 of 7 Archive leaderboard report
Overlapped 15-1 PASCAL VOC 2012 PLOP mIoU 54.64 #9 of 13 Archive leaderboard report
Overlapped 15-1 PASCAL VOC 2012 MiB mIoU 29.29 #11 of 13 Archive leaderboard report
Overlapped 15-5 PASCAL VOC 2012 PLOP Mean IoU (val) 70.09 #9 of 13 Archive leaderboard report
Overlapped 15-5 PASCAL VOC 2012 MiB Mean IoU (val) 70.08 #10 of 13 Archive leaderboard report
Overlapped 50-50 ADE20K PLOP mIoU 30.4 #4 of 7 Archive leaderboard report
Overlapped 50-50 ADE20K MiB mIoU 29.31 #7 of 7 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections