Low-contrast medical image segmentation via transformer and boundary perception

Keyphrases

• Boundary Location 14%
• Boundary Perception 100%
• Challenging Tasks 14%
• Ciliary Body 14%
• Convolutional Neural Network 14%
• Corneal Endothelial Cells 14%
• Embedded Boundary 42%
• Embedded Module 14%
• Enhanced Transformer 28%
• Feature Learning 14%
• Feature Map 14%
• Field Weight 14%
• Full Use 14%
• Global Context 14%
• Global Information 14%
• Global Relation 14%
• Global-local 14%
• Human Segmentation 14%
• Kidney Segmentation 14%
• Local Details 57%
• Local Use 14%
• Local Weight 14%
• Long-range Dependence 42%
• Low Contrast 100%
• Medical Image Segmentation 100%
• Multi-head 14%
• Parallel Mode 14%
• Perception Network 42%
• Receptive Field 14%
• Robust Features 14%
• Segmentation Performance 14%
• Segmentation Problem 14%
• Self-attention 14%
• Semantic Context 14%
• State-of-the-art Techniques 14%
• Transformer 100%
• Weight Sharing 14%

Computer Science

• Contrast Condition 33%
• Convolutional Neural Network 33%
• Desirable Property 33%
• Global Context 33%
• Global Information 33%
• Image Segmentation 100%
• Local Information 33%
• Range Dependency 100%
• Receptive Field 33%
• Representation Learning 33%
• Segmentation Performance 33%
• Segmentation Task 33%
• Target Location 33%

Biochemistry, Genetics and Molecular Biology

• Receptive Field 100%