Chapter 43: Learned Reconstruction

Learning Objectives

• Implement MF→U-Net post-processing networks for RF imaging
• Implement unrolled ISTA and unrolled OAMP with learnable parameters
• Build PnP reconstruction using pre-trained DnCNN/DRUNet denoisers
• Implement diffusion posterior sampling (DPS) for RF image reconstruction
• Run end-to-end comparison of all methods on standardized benchmarks