A multi-purpose image forensic method using densely connected convolutional neural networks

Keyphrases

• Forensic 100%
• Image Manipulation 100%
• Image Forensics 100%
• JPEG Compression 100%
• Forensic Techniques 100%
• Densely Connected Convolutional Networks 100%
• Multiple Operations 50%
• Image Size 50%
• Pooling Layer 50%
• Convolutional Neural Network Architecture 50%
• State-of-the-art Techniques 50%
• Convolution Kernel 50%
• Convolutional Layer 50%
• Connectivity Pattern 50%
• Simultaneous Detection 50%
• Local Image 50%
• Four States 50%
• Small Images 50%
• Image Patch 50%
• Neural Network Structure 50%
• Operation Detection 50%
• Traditional Pattern 50%
• Efficient Convolutional Neural Network 50%
• Nonlinear Activation Function 50%
• Dense Connectivity 50%
• Image Manipulation Detection 50%
• Parameter Efficiency 50%

Computer Science

• Convolutional Neural Network 100%
• Forensic Image 100%
• Image Manipulation 75%
• Neural Network Architecture 50%
• Convolutional Layer 25%
• Related Feature 25%
• Architecture Component 25%