HyperAI

Semantic Segmentation On Lars

Metrics

F1
Q
mIoU
μ

Results

Performance results of various models on this benchmark

Model Name
F1
Q
mIoU
μ
Paper TitleRepository
BiSeNetv1 (ResNet-50)42.839.492.273.3LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark
STDC161.857.893.675.6LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark
STDC264.360.894.576.5LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark
BiSeNetv254.751.293.573.9LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark
SWIM^2 (Mask2Former)79.978.197.879.7 The 2nd Workshop on Maritime Computer Vision (MaCVi) 2024-
IntCatchAI44.920.545.662.4LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark
PointRend65.462.194.977.5LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark
DeepLabv3 (ResNet-101)66.162.995.277.5LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark
Segmenter (ViT-B)55.252.695.172.2LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark
WODIS (ResNet-101)47.540.785.763.0LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark
eWaSR58.956.596.067.8The 2nd Workshop on Maritime Computer Vision (MaCVi) 2024-
SegFormer (MiT-B2)70.067.896.878.6LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark
KNet (Swin-T)73.471.397.278.8LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark
UNet15.413.990.175.7LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark
TransMari (Mask2Former)80.277.897.179.6The 2nd Workshop on Maritime Computer Vision (MaCVi) 2024-
DeepLabv3+ (ResNet-101)64.061.095.477.8LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark
FCN (ResNet-50)57.953.692.676.8LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark
WaSR (ResNet-101)61.659.596.671.0LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark
Mari-Mask2Former (Mask2Former)77.375.797.878.4The 2nd Workshop on Maritime Computer Vision (MaCVi) 2024-
FCN (ResNet-101)63.460.295.077.4LaRS: A Diverse Panoptic Maritime Obstacle Detection Dataset and Benchmark
0 of 20 row(s) selected.