G-GANISR: Gradual generative adversarial network for image super resolution

Keyphrases

• Generative Adversarial Networks 100%
• Image Super-resolution 100%
• Learning Process 50%
• Real Image 50%
• Upsampling 50%
• Least Squares Loss 50%
• Scale Factor 25%
• Proposed Methodology 25%
• Learning Strategies 25%
• Training Process 25%
• Training Stability 25%
• Image Details 25%
• Super-resolution Image 25%
• Cross-entropy 25%
• Test Quality 25%
• Efficient Super-resolution 25%
• Training Difficulty 25%
• Residual Learning 25%
• Adversarial Process 25%
• Overall Stability 25%
• Dense Residual Set 25%
• Gradual Learning 25%

Computer Science

• super resolution 100%
• Generative Adversarial Networks 100%
• Discriminator 75%
• Least Squares Method 50%
• Learning Process 50%
• Objective Function 25%
• Resolution Image 25%
• Training Process 25%
• Scaling Factor 25%