英语课论文思维导图

英语课论文Nonhomogeneous Sea Clutter Suppression Using Complex-Valued U-Net Model思维导图，一起来看。

Nonhomogeneous Sea Clutter Suppression UsingComplex-Valued U-Net Model

• Topic
  • Sea clutter suppression
• Background
  • The performance of most traditional clutter suppression methods relies onclutter statistical models, the suppression performance will drop sharply when the real clutter is mismatched with the statistical model
  • Deep learning have been applied to clutter suppression and target detection without relying on a specific statistical model, but the existed methods ignore the phase of the signal, which may reduce suppression performance
• Proposed method
  • Sea clutter suppression and target detection architecture
• Experiments and analysis
  • Introduce the method of dataset construction and network training
  • The performance of the proposed clutter suppression and target detection architecture is evaluated via real sea data
  • The detection performance of CV-UNet and other methods is compared under different signal-to-clutter ratios (SCRs)
• Conclusion
  • The CV -UNet based clutter suppression method proposed can extract the phase information from the complex signals
  • The CV -UNet has better clutter suppression ability under complicated sea conditions, compared with RV -Unet
  • How to improve the generalization of the model to adapt different weather such as rainy days will be considered in the future