Learning from pseudo-lesion: a self-supervised framework for COVID-19 diagnosis

Computer Science

• Deep Learning 100%
• Diagnosis Task 100%
• Experimental Result 100%
• Image Restoration 100%
• Learning Approach 100%
• Self-Supervised Learning 100%
• U-Net 100%

Keyphrases

• Annotated Dataset 10%
• Architecture-centric 10%
• COIVD-19 10%
• Computed Tomography 10%
• Computed Tomography Images 20%
• COVID-19 100%
• Deep Learning Methods 10%
• Encoder 10%
• Encoder-decoder 10%
• Feature Representation 10%
• Gradient Noise 10%
• Ground-glass Opacity 10%
• Image Recognition 10%
• Image Reconstruction 10%
• Jinan 10%
• Labeled Data 20%
• Large Image 10%
• Lesion Generation 10%
• Lung Region 10%
• Mathematical Model 10%
• Perlin Noise 10%
• Pseudolesion 100%
• Recognition Task 10%
• Self-supervised Learning 10%
• Semi-supervised Training 10%
• Supervised Learning Method 10%
• Supervised Model 10%
• Thoracic Computed Tomography 10%
• U-Net 10%