Papers › LSKNet: A Foundation Lightweight Backbone for Remote Sensing

LSKNet: A Foundation Lightweight Backbone for Remote Sensing

18 Mar 2024arXiv:2403.11735archive 2025-07-28

YuXuan Li, Xiang Li, Yimian Dai, Qibin Hou, Li Liu, Yongxiang Liu, Ming-Ming Cheng, Jian Yang

Remote sensing images pose distinct challenges for downstream tasks due to their inherent complexity. While a considerable amount of research has been dedicated to remote sensing classification, object detection and semantic segmentation, most of these studies have overlooked the valuable prior knowledge embedded within remote sensing scenarios. Such prior knowledge can be useful because remote sensing objects may be mistakenly recognized without referencing a sufficiently long-range context, which can vary for different objects. This paper considers these priors and proposes a lightweight Large Selective Kernel Network (LSKNet) backbone. LSKNet can dynamically adjust its large spatial receptive field to better model the ranging context of various objects in remote sensing scenarios. To our knowledge, large and selective kernel mechanisms have not been previously explored in remote sensing images. Without bells and whistles, our lightweight LSKNet sets new state-of-the-art scores on standard remote sensing classification, object detection and semantic segmentation benchmarks. Our comprehensive analysis further validated the significance of the identified priors and the effectiveness of LSKNet. The code is available at https://github.com/zcablii/LSKNet.

PaperPDFCode

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

Code

zcablii/lsknet officialmentioned in papermentioned on GitHubpytorchNOASSERTION report
zcablii/Large-Selective-Kernel-Network mentioned on GitHubpytorchNOASSERTION 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

Change DetectionObject DetectionObject Detection In Aerial ImagesSemantic Segmentationobject-detection

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Change Detection LEVIR-CD LSKNet F1 92.27 #7 of 28 Archive leaderboard report
Change Detection LEVIR-CD LSKNet F1-score 92.27 #7 of 28 Archive leaderboard report
Change Detection LEVIR-CD LSKNet IoU 85.65 #7 of 28 Archive leaderboard report
Change Detection LEVIR-CD LSKNet Precision 93.34 #7 of 28 Archive leaderboard report
Change Detection LEVIR-CD LSKNet Recall 91.23 #7 of 28 Archive leaderboard report
Change Detection S2Looking LSKNet-S F1-Score 67.52 #4 of 11 Archive leaderboard report
Change Detection S2Looking LSKNet-S IoU 50.96 #4 of 11 Archive leaderboard report
Change Detection S2Looking LSKNet-S Precision 71.90 #4 of 11 Archive leaderboard report
Change Detection S2Looking LSKNet-S Recall 63.64 #4 of 11 Archive leaderboard report
Object Detection In Aerial Images DOTA LSKNet-S* mAP 81.85% #7 of 58 Archive leaderboard report
Object Detection In Aerial Images DOTA LSKNet-S mAP 81.64% #10 of 58 Archive leaderboard report
Object Detection In Aerial Images DOTA LSKNet-T mAP 81.37% #11 of 58 Archive leaderboard report
Semantic Segmentation ISPRS Potsdam LSKNet-S Mean F1 93.1 #5 of 20 Archive leaderboard report
Semantic Segmentation ISPRS Potsdam LSKNet-S Mean IoU 87.2 #5 of 20 Archive leaderboard report
Semantic Segmentation ISPRS Potsdam LSKNet-S Overall Accuracy 92.0 #5 of 20 Archive leaderboard report
Semantic Segmentation ISPRS Vaihingen LSKNet-S Average F1 91.8 #1 of 12 Archive leaderboard report
Semantic Segmentation ISPRS Vaihingen LSKNet-S Category mIoU 85.1 #1 of 12 Archive leaderboard report
Semantic Segmentation ISPRS Vaihingen LSKNet-S Overall Accuracy 93.6 #1 of 12 Archive leaderboard report
Semantic Segmentation ISPRS Vaihingen LSKNet-T Average F1 91.7 #2 of 12 Archive leaderboard report
Semantic Segmentation ISPRS Vaihingen LSKNet-T Category mIoU 84.9 #2 of 12 Archive leaderboard report
Semantic Segmentation ISPRS Vaihingen LSKNet-T Overall Accuracy 93.6 #2 of 12 Archive leaderboard report
Semantic Segmentation UAVid LSKNet-S Mean IoU 70.0 #5 of 10 Archive leaderboard report
Semantic Segmentation UAVid LSKNet-T Mean IoU 69.3 #6 of 10 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

1x1 ConvolutionBatch NormalizationDilated ConvolutionReLUSelective KernelSelective Kernel ConvolutionSoftmax

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