Deep convolutional neural networks meet variational shape compactness priors for image segmentation

Keyphrases

• Image Segmentation 100%
• Deep Convolutional Neural Network (deep CNN) 100%
• Variational 100%
• Threshold Dynamics 100%
• Shape Compactness 100%
• Deep Neural Network 60%
• Soft Threshold 60%
• Dynamic Algorithms 60%
• Circular Object 40%
• Prediction Accuracy 20%
• Segmentation Problem 20%
• State-of-the-art Techniques 20%
• Computational Efficiency 20%
• Non-convexity 20%
• Existing Algorithms 20%
• Original Problem 20%
• Inconsistency 20%
• Numerical Efficiency 20%
• Black Box 20%
• Image Segmentation Method 20%
• Optimization Algorithm 20%
• Hyperparameters 20%
• Adversarial Attack 20%
• Noise Level 20%
• Circularity 20%
• Iris Dataset 20%
• Numerical Effectiveness 20%
• Spatial Prior 20%
• Variational Segmentation 20%
• Noise Attacks 20%
• Sigmoid Function 20%
• Prediction Robustness 20%
• Inefficiency Model 20%

Computer Science

• Image Segmentation 100%
• Deep Convolutional Neural Networks 100%
• Deep Neural Network 75%
• Segmentation Task 25%
• Adversarial Machine Learning 25%
• Optimization Algorithm 25%
• Challenging Scenario 25%
• Leaning Parameter 25%
• Segmentation Method 25%
• Prediction Accuracy 25%
• Computational Efficiency 25%
• Sigmoid Activation Function 25%