SegmentMeIfYouCan: A Benchmark for Anomaly Segmentation.
Robin ChanKrzysztof LisSvenja UhlemeyerHermann BlumSina HonariRoland SiegwartPascal FuaMathieu SalzmannMatthias RottmannPublished in: NeurIPS Datasets and Benchmarks (2021)
Keyphrases
- segmentation algorithm
- level set
- berkeley segmentation dataset
- medical images
- region growing
- segmentation method
- anomaly detection
- image segmentation
- intrusion detection
- pairwise
- quantitative evaluation
- shape prior
- segmented images
- joint segmentation
- segmentation accuracy
- motion segmentation
- multiscale
- fully automatic
- image analysis
- automatic segmentation
- graph cuts
- outdoor scenes
- energy function
- word segmentation
- fully unsupervised
- segmentation errors
- brain mri
- piece wise
- optimal segmentation
- computer vision