Papers › Semi-Supervised Semantic Segmentation via Marginal Contextual Information

Semi-Supervised Semantic Segmentation via Marginal Contextual Information

26 Aug 2023arXiv:2308.13900archive 2025-07-28

Moshe Kimhi, Shai Kimhi, Evgenii Zheltonozhskii, Or Litany, Chaim Baskin

We present a novel confidence refinement scheme that enhances pseudo labels in semi-supervised semantic segmentation. Unlike existing methods, which filter pixels with low-confidence predictions in isolation, our approach leverages the spatial correlation of labels in segmentation maps by grouping neighboring pixels and considering their pseudo labels collectively. With this contextual information, our method, named S4MC, increases the amount of unlabeled data used during training while maintaining the quality of the pseudo labels, all with negligible computational overhead. Through extensive experiments on standard benchmarks, we demonstrate that S4MC outperforms existing state-of-the-art semi-supervised learning approaches, offering a promising solution for reducing the cost of acquiring dense annotations. For example, S4MC achieves a 1.39 mIoU improvement over the prior art on PASCAL VOC 12 with 366 annotated images. The code to reproduce our experiments is available at https://s4mcontext.github.io/

PaperPDFCode

Code

s4mcontext/s4mc officialpytorch 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

SegmentationSemantic SegmentationSemi-Supervised Semantic Segmentation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Semi-Supervised Semantic Segmentation COCO 1/128 labeled S4MC Validation mIoU 43.78 #7 of 9 Archive leaderboard report
Semi-Supervised Semantic Segmentation COCO 1/256 labeled S4MC Validation mIoU 40.4 #4 of 9 Archive leaderboard report
Semi-Supervised Semantic Segmentation COCO 1/32 labeled S4MC Validation mIoU 50.58 #3 of 7 Archive leaderboard report
Semi-Supervised Semantic Segmentation COCO 1/512 labeled S4MC Validation mIoU 32.9 #5 of 8 Archive leaderboard report
Semi-Supervised Semantic Segmentation COCO 1/64 labeled S4MC Validation mIoU 47.98 #7 of 9 Archive leaderboard report
Semi-Supervised Semantic Segmentation Cityscapes 12.5% labeled S4MC Validation mIoU 77.78% #10 of 33 Archive leaderboard report
Semi-Supervised Semantic Segmentation Cityscapes 25% labeled S4MC Validation mIoU 79.52% #5 of 30 Archive leaderboard report
Semi-Supervised Semantic Segmentation Cityscapes 50% labeled S4MC Validation mIoU 79.76% #9 of 23 Archive leaderboard report
Semi-Supervised Semantic Segmentation Cityscapes 6.25% labeled S4MC Validation mIoU 77.00 #6 of 18 Archive leaderboard report
Semi-Supervised Semantic Segmentation PASCAL VOC 2012 1464 labels S4MC Validation mIoU 81.56 #9 of 17 Archive leaderboard report
Semi-Supervised Semantic Segmentation PASCAL VOC 2012 1464 labels S4MC (R50) Validation mIoU 0.7941 #17 of 17 Archive leaderboard report
Semi-Supervised Semantic Segmentation PASCAL VOC 2012 183 labeled S4MC Validation mIoU 75.21 #13 of 16 Archive leaderboard report
Semi-Supervised Semantic Segmentation PASCAL VOC 2012 25% labeled S4MC Validation mIoU 79.85 #9 of 27 Archive leaderboard report
Semi-Supervised Semantic Segmentation PASCAL VOC 2012 366 labeled S4MC Validation mIoU 79.09 #9 of 15 Archive leaderboard report
Semi-Supervised Semantic Segmentation PASCAL VOC 2012 50% S4MC Validation mIoU 81.11 #1 of 14 Archive leaderboard report
Semi-Supervised Semantic Segmentation PASCAL VOC 2012 732 labeled S4MC Validation mIoU 80.12 #9 of 16 Archive leaderboard report
Semi-Supervised Semantic Segmentation PASCAL VOC 2012 732 labeled S4MC (R50) Validation mIoU 77.83 #13 of 16 Archive leaderboard report
Semi-Supervised Semantic Segmentation PASCAL VOC 2012 92 labeled S4MC Validation mIoU 74.72 #11 of 17 Archive leaderboard report
Semi-Supervised Semantic Segmentation Pascal VOC 2012 12.5% labeled S4MC Validation mIoU 79.67% #9 of 38 Archive leaderboard report
Semi-Supervised Semantic Segmentation Pascal VOC 2012 6.25% labeled S4MC Validation mIoU 78.84 #7 of 19 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