Papers › SATS: Self-Attention Transfer for Continual Semantic Segmentation

SATS: Self-Attention Transfer for Continual Semantic Segmentation

15 Mar 2022arXiv:2203.07667archive 2025-07-28

Yiqiao Qiu, Yixing Shen, Zhuohao Sun, Yanchong Zheng, Xiaobin Chang, Weishi Zheng, Ruixuan Wang

Continually learning to segment more and more types of image regions is a desired capability for many intelligent systems. However, such continual semantic segmentation suffers from the same catastrophic forgetting issue as in continual classification learning. While multiple knowledge distillation strategies originally for continual classification have been well adapted to continual semantic segmentation, they only consider transferring old knowledge based on the outputs from one or more layers of deep fully convolutional networks. Different from existing solutions, this study proposes to transfer a new type of information relevant to knowledge, i.e. the relationships between elements (Eg. pixels or small local regions) within each image which can capture both within-class and between-class knowledge. The relationship information can be effectively obtained from the self-attention maps in a Transformer-style segmentation model. Considering that pixels belonging to the same class in each image often share similar visual properties, a class-specific region pooling is applied to provide more efficient relationship information for knowledge transfer. Extensive evaluations on multiple public benchmarks support that the proposed self-attention transfer method can further effectively alleviate the catastrophic forgetting issue, and its flexible combination with one or more widely adopted strategies significantly outperforms state-of-the-art solutions.

PaperPDFCode

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

Code

QIU023/SATS_Continual_Semantic_Seg officialmentioned 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

Continual Semantic SegmentationKnowledge DistillationOverlapped 10-1Overlapped 100-10Overlapped 15-1Overlapped 15-5Overlapped 25-25Overlapped 5-3SegmentationSemantic SegmentationTransfer Learning

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Overlapped 10-1 PASCAL VOC 2012 SATS-M mIoU 69.27 #2 of 13 Archive leaderboard report
Overlapped 10-1 PASCAL VOC 2012 SATS mIoU 61.6 #4 of 13 Archive leaderboard report
Overlapped 100-10 ADE20K SATS-M Mean IoU (test) 35.45 #2 of 6 Archive leaderboard report
Overlapped 15-1 PASCAL VOC 2012 SATS-M mIoU 76.61 #2 of 13 Archive leaderboard report
Overlapped 15-1 PASCAL VOC 2012 SATS mIoU 74.48 #3 of 13 Archive leaderboard report
Overlapped 15-5 PASCAL VOC 2012 SATS-M Mean IoU (val) 78.72 #2 of 13 Archive leaderboard report
Overlapped 15-5 PASCAL VOC 2012 SATS Mean IoU (val) 75.70 #3 of 13 Archive leaderboard report
Overlapped 25-25 ADE20K SATS-M Mean IoU (test) 32.56 #1 of 1 Archive leaderboard report
Overlapped 5-3 PASCAL VOC 2012 SATS-M Mean IoU (test) 71.36 #2 of 4 Archive leaderboard report
Overlapped 5-3 PASCAL VOC 2012 SATS Mean IoU (test) 67.36 #3 of 4 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.

Methods

Knowledge Distillation

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