GAN-SNR-Shrinkage-Based Network for Modulation Recognition with Small Training Sample Size

Computer Science

• Training Sample 100%
• Noise-to-Signal Ratio 100%
• Generative Adversarial Networks 100%
• Cooperative Communication 50%
• Training Data 50%
• Few-Shot Learning 50%
• Training Dataset 50%
• Baseline Method 50%
• Preprocessing Module 50%
• Recognition Network 50%
• Play Recognition 50%
• Noise Performance 50%

Engineering

• Signal-to-Noise Ratio 100%
• Cooperative 50%
• Noise Performance 50%
• Residual Shrinkage 50%