TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Papers/LaRS: A Diverse Panoptic Maritime Obstacle Detection Datas...

LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark

Lojze Žust, Janez Perš, Matej Kristan

2023-08-18ICCV 2023 1Panoptic SegmentationSemantic SegmentationVideo Semantic Segmentation
PaperPDFCodeCode(official)

Abstract

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

Results

TaskDatasetMetricValueModel
Scene ParsingLaRSF162.1WaSR-T (ResNet-101)
Scene ParsingLaRSQ60.1WaSR-T (ResNet-101)
Scene ParsingLaRSmIoU96.7WaSR-T (ResNet-101)
Scene ParsingLaRSμ71.1WaSR-T (ResNet-101)
Scene ParsingLaRSF161.1TMANet (ResNet-50)
Scene ParsingLaRSQ57.5TMANet (ResNet-50)
Scene ParsingLaRSmIoU94.1TMANet (ResNet-50)
Scene ParsingLaRSμ77.1TMANet (ResNet-50)
Scene ParsingLaRSF152.1CSANet (ResNet-101)
Scene ParsingLaRSQ49.1CSANet (ResNet-101)
Scene ParsingLaRSmIoU94.2CSANet (ResNet-101)
Scene ParsingLaRSμ63.7CSANet (ResNet-101)
Semantic SegmentationLaRSF173.4KNet (Swin-T)
Semantic SegmentationLaRSQ71.3KNet (Swin-T)
Semantic SegmentationLaRSmIoU97.2KNet (Swin-T)
Semantic SegmentationLaRSμ78.8KNet (Swin-T)
Semantic SegmentationLaRSF170SegFormer (MiT-B2)
Semantic SegmentationLaRSQ67.8SegFormer (MiT-B2)
Semantic SegmentationLaRSmIoU96.8SegFormer (MiT-B2)
Semantic SegmentationLaRSμ78.6SegFormer (MiT-B2)
Semantic SegmentationLaRSF166.1DeepLabv3 (ResNet-101)
Semantic SegmentationLaRSQ62.9DeepLabv3 (ResNet-101)
Semantic SegmentationLaRSmIoU95.2DeepLabv3 (ResNet-101)
Semantic SegmentationLaRSμ77.5DeepLabv3 (ResNet-101)
Semantic SegmentationLaRSF165.4PointRend
Semantic SegmentationLaRSQ62.1PointRend
Semantic SegmentationLaRSmIoU94.9PointRend
Semantic SegmentationLaRSμ77.5PointRend
Semantic SegmentationLaRSF164DeepLabv3+ (ResNet-101)
Semantic SegmentationLaRSQ61DeepLabv3+ (ResNet-101)
Semantic SegmentationLaRSmIoU95.4DeepLabv3+ (ResNet-101)
Semantic SegmentationLaRSμ77.8DeepLabv3+ (ResNet-101)
Semantic SegmentationLaRSF164.3STDC2
Semantic SegmentationLaRSQ60.8STDC2
Semantic SegmentationLaRSmIoU94.5STDC2
Semantic SegmentationLaRSμ76.5STDC2
Semantic SegmentationLaRSF163.4FCN (ResNet-101)
Semantic SegmentationLaRSQ60.2FCN (ResNet-101)
Semantic SegmentationLaRSmIoU95FCN (ResNet-101)
Semantic SegmentationLaRSμ77.4FCN (ResNet-101)
Semantic SegmentationLaRSF161.6WaSR (ResNet-101)
Semantic SegmentationLaRSQ59.5WaSR (ResNet-101)
Semantic SegmentationLaRSmIoU96.6WaSR (ResNet-101)
Semantic SegmentationLaRSμ71WaSR (ResNet-101)
Semantic SegmentationLaRSF161.8STDC1
Semantic SegmentationLaRSQ57.8STDC1
Semantic SegmentationLaRSmIoU93.6STDC1
Semantic SegmentationLaRSμ75.6STDC1
Semantic SegmentationLaRSF157.9FCN (ResNet-50)
Semantic SegmentationLaRSQ53.6FCN (ResNet-50)
Semantic SegmentationLaRSmIoU92.6FCN (ResNet-50)
Semantic SegmentationLaRSμ76.8FCN (ResNet-50)
Semantic SegmentationLaRSF155.2Segmenter (ViT-B)
Semantic SegmentationLaRSQ52.6Segmenter (ViT-B)
Semantic SegmentationLaRSmIoU95.1Segmenter (ViT-B)
Semantic SegmentationLaRSμ72.2Segmenter (ViT-B)
Semantic SegmentationLaRSF154.7BiSeNetv2
Semantic SegmentationLaRSQ51.2BiSeNetv2
Semantic SegmentationLaRSmIoU93.5BiSeNetv2
Semantic SegmentationLaRSμ73.9BiSeNetv2
Semantic SegmentationLaRSF147.5WODIS (ResNet-101)
Semantic SegmentationLaRSQ40.7WODIS (ResNet-101)
Semantic SegmentationLaRSmIoU85.7WODIS (ResNet-101)
Semantic SegmentationLaRSμ63WODIS (ResNet-101)
Semantic SegmentationLaRSF142.8BiSeNetv1 (ResNet-50)
Semantic SegmentationLaRSQ39.4BiSeNetv1 (ResNet-50)
Semantic SegmentationLaRSmIoU92.2BiSeNetv1 (ResNet-50)
Semantic SegmentationLaRSμ73.3BiSeNetv1 (ResNet-50)
Semantic SegmentationLaRSF144.9IntCatchAI
Semantic SegmentationLaRSQ20.5IntCatchAI
Semantic SegmentationLaRSmIoU45.6IntCatchAI
Semantic SegmentationLaRSμ62.4IntCatchAI
Semantic SegmentationLaRSF115.4UNet
Semantic SegmentationLaRSQ13.9UNet
Semantic SegmentationLaRSmIoU90.1UNet
Semantic SegmentationLaRSμ75.7UNet
Semantic SegmentationLaRSPQ41.7Mask2Former (Swin-B)
Semantic SegmentationLaRSPQ40.1Panoptic FPN (ResNet-50)
Semantic SegmentationLaRSPQ39.2Mask2Former (Swin-T)
Semantic SegmentationLaRSPQ38.7Panoptic FPN (ResNet-101)
Semantic SegmentationLaRSPQ37.6Mask2Former (ResNet-50)
Semantic SegmentationLaRSPQ37.2Mask2Former (ResNet-101)
Semantic SegmentationLaRSPQ34.7Panoptic Deeplab (ResNet-50)
Semantic SegmentationLaRSPQ31.9MaX-DeepLab
Video Semantic SegmentationLaRSF162.1WaSR-T (ResNet-101)
Video Semantic SegmentationLaRSQ60.1WaSR-T (ResNet-101)
Video Semantic SegmentationLaRSmIoU96.7WaSR-T (ResNet-101)
Video Semantic SegmentationLaRSμ71.1WaSR-T (ResNet-101)
Video Semantic SegmentationLaRSF161.1TMANet (ResNet-50)
Video Semantic SegmentationLaRSQ57.5TMANet (ResNet-50)
Video Semantic SegmentationLaRSmIoU94.1TMANet (ResNet-50)
Video Semantic SegmentationLaRSμ77.1TMANet (ResNet-50)
Video Semantic SegmentationLaRSF152.1CSANet (ResNet-101)
Video Semantic SegmentationLaRSQ49.1CSANet (ResNet-101)
Video Semantic SegmentationLaRSmIoU94.2CSANet (ResNet-101)
Video Semantic SegmentationLaRSμ63.7CSANet (ResNet-101)
Scene UnderstandingLaRSF162.1WaSR-T (ResNet-101)
Scene UnderstandingLaRSQ60.1WaSR-T (ResNet-101)
Scene UnderstandingLaRSmIoU96.7WaSR-T (ResNet-101)
Scene UnderstandingLaRSμ71.1WaSR-T (ResNet-101)
Scene UnderstandingLaRSF161.1TMANet (ResNet-50)
Scene UnderstandingLaRSQ57.5TMANet (ResNet-50)
Scene UnderstandingLaRSmIoU94.1TMANet (ResNet-50)
Scene UnderstandingLaRSμ77.1TMANet (ResNet-50)
Scene UnderstandingLaRSF152.1CSANet (ResNet-101)
Scene UnderstandingLaRSQ49.1CSANet (ResNet-101)
Scene UnderstandingLaRSmIoU94.2CSANet (ResNet-101)
Scene UnderstandingLaRSμ63.7CSANet (ResNet-101)
2D Semantic SegmentationLaRSF162.1WaSR-T (ResNet-101)
2D Semantic SegmentationLaRSQ60.1WaSR-T (ResNet-101)
2D Semantic SegmentationLaRSmIoU96.7WaSR-T (ResNet-101)
2D Semantic SegmentationLaRSμ71.1WaSR-T (ResNet-101)
2D Semantic SegmentationLaRSF161.1TMANet (ResNet-50)
2D Semantic SegmentationLaRSQ57.5TMANet (ResNet-50)
2D Semantic SegmentationLaRSmIoU94.1TMANet (ResNet-50)
2D Semantic SegmentationLaRSμ77.1TMANet (ResNet-50)
2D Semantic SegmentationLaRSF152.1CSANet (ResNet-101)
2D Semantic SegmentationLaRSQ49.1CSANet (ResNet-101)
2D Semantic SegmentationLaRSmIoU94.2CSANet (ResNet-101)
2D Semantic SegmentationLaRSμ63.7CSANet (ResNet-101)
10-shot image generationLaRSF173.4KNet (Swin-T)
10-shot image generationLaRSQ71.3KNet (Swin-T)
10-shot image generationLaRSmIoU97.2KNet (Swin-T)
10-shot image generationLaRSμ78.8KNet (Swin-T)
10-shot image generationLaRSF170SegFormer (MiT-B2)
10-shot image generationLaRSQ67.8SegFormer (MiT-B2)
10-shot image generationLaRSmIoU96.8SegFormer (MiT-B2)
10-shot image generationLaRSμ78.6SegFormer (MiT-B2)
10-shot image generationLaRSF166.1DeepLabv3 (ResNet-101)
10-shot image generationLaRSQ62.9DeepLabv3 (ResNet-101)
10-shot image generationLaRSmIoU95.2DeepLabv3 (ResNet-101)
10-shot image generationLaRSμ77.5DeepLabv3 (ResNet-101)
10-shot image generationLaRSF165.4PointRend
10-shot image generationLaRSQ62.1PointRend
10-shot image generationLaRSmIoU94.9PointRend
10-shot image generationLaRSμ77.5PointRend
10-shot image generationLaRSF164DeepLabv3+ (ResNet-101)
10-shot image generationLaRSQ61DeepLabv3+ (ResNet-101)
10-shot image generationLaRSmIoU95.4DeepLabv3+ (ResNet-101)
10-shot image generationLaRSμ77.8DeepLabv3+ (ResNet-101)
10-shot image generationLaRSF164.3STDC2
10-shot image generationLaRSQ60.8STDC2
10-shot image generationLaRSmIoU94.5STDC2
10-shot image generationLaRSμ76.5STDC2
10-shot image generationLaRSF163.4FCN (ResNet-101)
10-shot image generationLaRSQ60.2FCN (ResNet-101)
10-shot image generationLaRSmIoU95FCN (ResNet-101)
10-shot image generationLaRSμ77.4FCN (ResNet-101)
10-shot image generationLaRSF161.6WaSR (ResNet-101)
10-shot image generationLaRSQ59.5WaSR (ResNet-101)
10-shot image generationLaRSmIoU96.6WaSR (ResNet-101)
10-shot image generationLaRSμ71WaSR (ResNet-101)
10-shot image generationLaRSF161.8STDC1
10-shot image generationLaRSQ57.8STDC1
10-shot image generationLaRSmIoU93.6STDC1
10-shot image generationLaRSμ75.6STDC1
10-shot image generationLaRSF157.9FCN (ResNet-50)
10-shot image generationLaRSQ53.6FCN (ResNet-50)
10-shot image generationLaRSmIoU92.6FCN (ResNet-50)
10-shot image generationLaRSμ76.8FCN (ResNet-50)
10-shot image generationLaRSF155.2Segmenter (ViT-B)
10-shot image generationLaRSQ52.6Segmenter (ViT-B)
10-shot image generationLaRSmIoU95.1Segmenter (ViT-B)
10-shot image generationLaRSμ72.2Segmenter (ViT-B)
10-shot image generationLaRSF154.7BiSeNetv2
10-shot image generationLaRSQ51.2BiSeNetv2
10-shot image generationLaRSmIoU93.5BiSeNetv2
10-shot image generationLaRSμ73.9BiSeNetv2
10-shot image generationLaRSF147.5WODIS (ResNet-101)
10-shot image generationLaRSQ40.7WODIS (ResNet-101)
10-shot image generationLaRSmIoU85.7WODIS (ResNet-101)
10-shot image generationLaRSμ63WODIS (ResNet-101)
10-shot image generationLaRSF142.8BiSeNetv1 (ResNet-50)
10-shot image generationLaRSQ39.4BiSeNetv1 (ResNet-50)
10-shot image generationLaRSmIoU92.2BiSeNetv1 (ResNet-50)
10-shot image generationLaRSμ73.3BiSeNetv1 (ResNet-50)
10-shot image generationLaRSF144.9IntCatchAI
10-shot image generationLaRSQ20.5IntCatchAI
10-shot image generationLaRSmIoU45.6IntCatchAI
10-shot image generationLaRSμ62.4IntCatchAI
10-shot image generationLaRSF115.4UNet
10-shot image generationLaRSQ13.9UNet
10-shot image generationLaRSmIoU90.1UNet
10-shot image generationLaRSμ75.7UNet
10-shot image generationLaRSPQ41.7Mask2Former (Swin-B)
10-shot image generationLaRSPQ40.1Panoptic FPN (ResNet-50)
10-shot image generationLaRSPQ39.2Mask2Former (Swin-T)
10-shot image generationLaRSPQ38.7Panoptic FPN (ResNet-101)
10-shot image generationLaRSPQ37.6Mask2Former (ResNet-50)
10-shot image generationLaRSPQ37.2Mask2Former (ResNet-101)
10-shot image generationLaRSPQ34.7Panoptic Deeplab (ResNet-50)
10-shot image generationLaRSPQ31.9MaX-DeepLab
Panoptic SegmentationLaRSPQ41.7Mask2Former (Swin-B)
Panoptic SegmentationLaRSPQ40.1Panoptic FPN (ResNet-50)
Panoptic SegmentationLaRSPQ39.2Mask2Former (Swin-T)
Panoptic SegmentationLaRSPQ38.7Panoptic FPN (ResNet-101)
Panoptic SegmentationLaRSPQ37.6Mask2Former (ResNet-50)
Panoptic SegmentationLaRSPQ37.2Mask2Former (ResNet-101)
Panoptic SegmentationLaRSPQ34.7Panoptic Deeplab (ResNet-50)
Panoptic SegmentationLaRSPQ31.9MaX-DeepLab

Related Papers

SeC: Advancing Complex Video Object Segmentation via Progressive Concept Construction2025-07-21DiffOSeg: Omni Medical Image Segmentation via Multi-Expert Collaboration Diffusion Model2025-07-17SCORE: Scene Context Matters in Open-Vocabulary Remote Sensing Instance Segmentation2025-07-17Unified Medical Image Segmentation with State Space Modeling Snake2025-07-17A Privacy-Preserving Semantic-Segmentation Method Using Domain-Adaptation Technique2025-07-17SAMST: A Transformer framework based on SAM pseudo label filtering for remote sensing semi-supervised semantic segmentation2025-07-16Tomato Multi-Angle Multi-Pose Dataset for Fine-Grained Phenotyping2025-07-15U-RWKV: Lightweight medical image segmentation with direction-adaptive RWKV2025-07-15