Factoring 3D Convolutions for Medical Images by Depth-wise Dependencies-induced Adaptive Attention

Keyphrases

• Medical Image 100%
• Convolutional Neural Network 100%
• Convolution Kernel 100%
• 3D Convolution 100%
• Adaptive Attention 100%
• Model Complexity 66%
• Learnability 66%
• Weight Matrix 66%
• Spectral Normalization 66%
• Factorized Convolution 66%
• Attention Block 66%
• 3D Medical Images 66%
• Computer-aided Diagnosis (CADx) 33%
• Model Performance 33%
• Depth Information 33%
• Hyperparameters 33%
• Medical Image Processing 33%
• Matrix Factorization 33%
• Information Processing Capability 33%
• Lung CT 33%
• Spectral Norm 33%
• Computing Power 33%
• Limited Computing Resources 33%
• Cube Structures 33%
• Computing Model 33%
• Depth Direction 33%
• LUng Nodule Analysis 2016 (LUNA16) 33%
• Resource Complexity 33%
• CT Dataset 33%
• Retina OCT 33%

Computer Science

• Convolutional Neural Network 100%
• Model Complexity 66%
• Learning Ability 66%
• Aided Diagnosis 33%
• Performance Model 33%
• Depth Information 33%
• Matrix Factorization 33%
• Convolution Kernel 33%
• Medical Image Processing 33%
• Processing Capability 33%
• Computing Resource 33%
• Computing Power 33%

Mathematics

• Convolution 100%
• Convolutional Neural Network 75%
• Weight Matrix 50%
• Matrix (Mathematics) 25%
• Performance Model 25%
• Constrains 25%
• Cube 25%
• Factorization 25%
• Medical Image Processing 25%