Papers › LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark

LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark

18 Aug 2023ICCV 2023 1arXiv:2308.09618archive 2025-07-28

Lojze Žust, Janez Perš, Matej Kristan

The progress in maritime obstacle detection is hindered by the lack of a diverse dataset that adequately captures the complexity of general maritime environments. We present the first maritime panoptic obstacle detection benchmark LaRS, featuring scenes from Lakes, Rivers and Seas. Our major contribution is the new dataset, which boasts the largest diversity in recording locations, scene types, obstacle classes, and acquisition conditions among the related datasets. LaRS is composed of over 4000 per-pixel labeled key frames with nine preceding frames to allow utilization of the temporal texture, amounting to over 40k frames. Each key frame is annotated with 8 thing, 3 stuff classes and 19 global scene attributes. We report the results of 27 semantic and panoptic segmentation methods, along with several performance insights and future research directions. To enable objective evaluation, we have implemented an online evaluation server. The LaRS dataset, evaluation toolkit and benchmark are publicly available at: https://lojzezust.github.io/lars-dataset

PaperPDFConference PDFCode

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

Code

lojzezust/lars_evaluator mentioned on GitHub 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

DiversityPanoptic SegmentationSemantic SegmentationVideo Semantic Segmentation

Datasets

Introduced by this paper, per the archive.

LaRS

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Panoptic Segmentation LaRS Mask2Former (Swin-B) PQ 41.7 #1 of 8 Archive leaderboard report
Panoptic Segmentation LaRS Panoptic FPN (ResNet-50) PQ 40.1 #2 of 8 Archive leaderboard report
Panoptic Segmentation LaRS Mask2Former (Swin-T) PQ 39.2 #3 of 8 Archive leaderboard report
Panoptic Segmentation LaRS Panoptic FPN (ResNet-101) PQ 38.7 #4 of 8 Archive leaderboard report
Panoptic Segmentation LaRS Mask2Former (ResNet-50) PQ 37.6 #5 of 8 Archive leaderboard report
Panoptic Segmentation LaRS Mask2Former (ResNet-101) PQ 37.2 #6 of 8 Archive leaderboard report
Panoptic Segmentation LaRS Panoptic Deeplab (ResNet-50) PQ 34.7 #7 of 8 Archive leaderboard report
Panoptic Segmentation LaRS MaX-DeepLab PQ 31.9 #8 of 8 Archive leaderboard report
Semantic Segmentation LaRS KNet (Swin-T) F1 73.4 #4 of 20 Archive leaderboard report
Semantic Segmentation LaRS KNet (Swin-T) Q 71.3 #4 of 20 Archive leaderboard report
Semantic Segmentation LaRS KNet (Swin-T) mIoU 97.2 #4 of 20 Archive leaderboard report
Semantic Segmentation LaRS KNet (Swin-T) μ 78.8 #4 of 20 Archive leaderboard report
Semantic Segmentation LaRS SegFormer (MiT-B2) F1 70.0 #5 of 20 Archive leaderboard report
Semantic Segmentation LaRS SegFormer (MiT-B2) Q 67.8 #5 of 20 Archive leaderboard report
Semantic Segmentation LaRS SegFormer (MiT-B2) mIoU 96.8 #5 of 20 Archive leaderboard report
Semantic Segmentation LaRS SegFormer (MiT-B2) μ 78.6 #5 of 20 Archive leaderboard report
Semantic Segmentation LaRS DeepLabv3 (ResNet-101) F1 66.1 #6 of 20 Archive leaderboard report
Semantic Segmentation LaRS DeepLabv3 (ResNet-101) Q 62.9 #6 of 20 Archive leaderboard report
Semantic Segmentation LaRS DeepLabv3 (ResNet-101) mIoU 95.2 #6 of 20 Archive leaderboard report
Semantic Segmentation LaRS DeepLabv3 (ResNet-101) μ 77.5 #6 of 20 Archive leaderboard report
Semantic Segmentation LaRS PointRend F1 65.4 #7 of 20 Archive leaderboard report
Semantic Segmentation LaRS PointRend Q 62.1 #7 of 20 Archive leaderboard report
Semantic Segmentation LaRS PointRend mIoU 94.9 #7 of 20 Archive leaderboard report
Semantic Segmentation LaRS PointRend μ 77.5 #7 of 20 Archive leaderboard report
Semantic Segmentation LaRS DeepLabv3+ (ResNet-101) F1 64.0 #8 of 20 Archive leaderboard report
Semantic Segmentation LaRS DeepLabv3+ (ResNet-101) Q 61.0 #8 of 20 Archive leaderboard report
Semantic Segmentation LaRS DeepLabv3+ (ResNet-101) mIoU 95.4 #8 of 20 Archive leaderboard report
Semantic Segmentation LaRS DeepLabv3+ (ResNet-101) μ 77.8 #8 of 20 Archive leaderboard report
Semantic Segmentation LaRS STDC2 F1 64.3 #9 of 20 Archive leaderboard report
Semantic Segmentation LaRS STDC2 Q 60.8 #9 of 20 Archive leaderboard report
Semantic Segmentation LaRS STDC2 mIoU 94.5 #9 of 20 Archive leaderboard report
Semantic Segmentation LaRS STDC2 μ 76.5 #9 of 20 Archive leaderboard report
Semantic Segmentation LaRS FCN (ResNet-101) F1 63.4 #10 of 20 Archive leaderboard report
Semantic Segmentation LaRS FCN (ResNet-101) Q 60.2 #10 of 20 Archive leaderboard report
Semantic Segmentation LaRS FCN (ResNet-101) mIoU 95.0 #10 of 20 Archive leaderboard report
Semantic Segmentation LaRS FCN (ResNet-101) μ 77.4 #10 of 20 Archive leaderboard report
Semantic Segmentation LaRS WaSR (ResNet-101) F1 61.6 #11 of 20 Archive leaderboard report
Semantic Segmentation LaRS WaSR (ResNet-101) Q 59.5 #11 of 20 Archive leaderboard report
Semantic Segmentation LaRS WaSR (ResNet-101) mIoU 96.6 #11 of 20 Archive leaderboard report
Semantic Segmentation LaRS WaSR (ResNet-101) μ 71.0 #11 of 20 Archive leaderboard report
Semantic Segmentation LaRS STDC1 F1 61.8 #12 of 20 Archive leaderboard report
Semantic Segmentation LaRS STDC1 Q 57.8 #12 of 20 Archive leaderboard report
Semantic Segmentation LaRS STDC1 mIoU 93.6 #12 of 20 Archive leaderboard report
Semantic Segmentation LaRS STDC1 μ 75.6 #12 of 20 Archive leaderboard report
Semantic Segmentation LaRS FCN (ResNet-50) F1 57.9 #14 of 20 Archive leaderboard report
Semantic Segmentation LaRS FCN (ResNet-50) Q 53.6 #14 of 20 Archive leaderboard report
Semantic Segmentation LaRS FCN (ResNet-50) mIoU 92.6 #14 of 20 Archive leaderboard report
Semantic Segmentation LaRS FCN (ResNet-50) μ 76.8 #14 of 20 Archive leaderboard report
Semantic Segmentation LaRS Segmenter (ViT-B) F1 55.2 #15 of 20 Archive leaderboard report
Semantic Segmentation LaRS Segmenter (ViT-B) Q 52.6 #15 of 20 Archive leaderboard report
Semantic Segmentation LaRS Segmenter (ViT-B) mIoU 95.1 #15 of 20 Archive leaderboard report
Semantic Segmentation LaRS Segmenter (ViT-B) μ 72.2 #15 of 20 Archive leaderboard report
Semantic Segmentation LaRS BiSeNetv2 F1 54.7 #16 of 20 Archive leaderboard report
Semantic Segmentation LaRS BiSeNetv2 Q 51.2 #16 of 20 Archive leaderboard report
Semantic Segmentation LaRS BiSeNetv2 mIoU 93.5 #16 of 20 Archive leaderboard report
Semantic Segmentation LaRS BiSeNetv2 μ 73.9 #16 of 20 Archive leaderboard report
Semantic Segmentation LaRS WODIS (ResNet-101) F1 47.5 #17 of 20 Archive leaderboard report
Semantic Segmentation LaRS WODIS (ResNet-101) Q 40.7 #17 of 20 Archive leaderboard report
Semantic Segmentation LaRS WODIS (ResNet-101) mIoU 85.7 #17 of 20 Archive leaderboard report
Semantic Segmentation LaRS WODIS (ResNet-101) μ 63.0 #17 of 20 Archive leaderboard report
Semantic Segmentation LaRS BiSeNetv1 (ResNet-50) F1 42.8 #18 of 20 Archive leaderboard report
Semantic Segmentation LaRS BiSeNetv1 (ResNet-50) Q 39.4 #18 of 20 Archive leaderboard report
Semantic Segmentation LaRS BiSeNetv1 (ResNet-50) mIoU 92.2 #18 of 20 Archive leaderboard report
Semantic Segmentation LaRS BiSeNetv1 (ResNet-50) μ 73.3 #18 of 20 Archive leaderboard report
Semantic Segmentation LaRS IntCatchAI F1 44.9 #19 of 20 Archive leaderboard report
Semantic Segmentation LaRS IntCatchAI Q 20.5 #19 of 20 Archive leaderboard report
Semantic Segmentation LaRS IntCatchAI mIoU 45.6 #19 of 20 Archive leaderboard report
Semantic Segmentation LaRS IntCatchAI μ 62.4 #19 of 20 Archive leaderboard report
Semantic Segmentation LaRS UNet F1 15.4 #20 of 20 Archive leaderboard report
Semantic Segmentation LaRS UNet Q 13.9 #20 of 20 Archive leaderboard report
Semantic Segmentation LaRS UNet mIoU 90.1 #20 of 20 Archive leaderboard report
Semantic Segmentation LaRS UNet μ 75.7 #20 of 20 Archive leaderboard report
Video Semantic Segmentation LaRS WaSR-T (ResNet-101) F1 62.1 #1 of 3 Archive leaderboard report
Video Semantic Segmentation LaRS WaSR-T (ResNet-101) Q 60.1 #1 of 3 Archive leaderboard report
Video Semantic Segmentation LaRS WaSR-T (ResNet-101) mIoU 96.7 #1 of 3 Archive leaderboard report
Video Semantic Segmentation LaRS WaSR-T (ResNet-101) μ 71.1 #1 of 3 Archive leaderboard report
Video Semantic Segmentation LaRS TMANet (ResNet-50) F1 61.1 #2 of 3 Archive leaderboard report
Video Semantic Segmentation LaRS TMANet (ResNet-50) Q 57.5 #2 of 3 Archive leaderboard report
Video Semantic Segmentation LaRS TMANet (ResNet-50) mIoU 94.1 #2 of 3 Archive leaderboard report
Video Semantic Segmentation LaRS TMANet (ResNet-50) μ 77.1 #2 of 3 Archive leaderboard report
Video Semantic Segmentation LaRS CSANet (ResNet-101) F1 52.1 #3 of 3 Archive leaderboard report
Video Semantic Segmentation LaRS CSANet (ResNet-101) Q 49.1 #3 of 3 Archive leaderboard report
Video Semantic Segmentation LaRS CSANet (ResNet-101) mIoU 94.2 #3 of 3 Archive leaderboard report
Video Semantic Segmentation LaRS CSANet (ResNet-101) μ 63.7 #3 of 3 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