Pixel-Level and Perceptual-Level Regularized Adversarial Learning for Joint Motion Deblurring and Super-Resolution

Keyphrases

• Super-resolution 100%
• Joint Motion 100%
• Pixel Level 100%
• Adversarial Learning 100%
• Motion Deblurring 100%
• Perceptual Level 100%
• Residual in Residual 100%
• Deblurring 80%
• Residual Dense 60%
• Contextual Loss 40%
• Adversarial Loss 40%
• Channel Attention Block 40%
• High-resolution 20%
• Clear Image 20%
• High-frequency Details 20%
• Feature Extracting 20%
• Pixel-wise 20%
• Dense Structure 20%
• Memory Mechanism 20%
• Regularization Term 20%
• Publicly Available Dataset 20%
• Encoder-decoder Architecture 20%
• Channel Features 20%
• Motion Blur 20%
• Blur Kernel 20%
• End-to-end Deep Learning 20%
• Channel-wise 20%
• Convolutional Filter 20%
• Receptive Field 20%
• Residual Structure 20%
• Motion-blurred Image 20%
• Continuous Memory 20%

Computer Science

• super resolution 100%
• Regularization Term 33%
• Convolutional Filter 33%
• Neural Network 33%
• Extracted Feature 33%
• Receptive Field 33%