Research map / Image generation / Autoregressive and latent generation
A Unified Framework for U-Net Design and Analysis
Christopher Williams, Fabian Falck, George Deligiannidis, Christopher C. Holmes, Arnaud Doucet, Saifuddin Syed
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace learned encoder with parameter-free Haar DWT encoder and move saved parameters into decoder (Multi-ResNet) | Residual U-Net · Navier-Stokes | r-MSE: 0.0057 -> 0.0040 | Improved |
| replace learned encoder with parameter-free Haar DWT encoder and move saved parameters into decoder (Multi-ResNet) | Residual U-Net · Shallow water | r-MSE: 0.1712 -> 0.1493 | Improved |
| replace learned encoder with parameter-free Haar DWT encoder and move saved parameters into decoder (Multi-ResNet) | Residual U-Net · WMH segmentation | Dice: 0.8069 -> 0.8346 | Improved |
| replace learned encoder with Haar DWT encoder and add saved parameters to decoder (Multi-ResNet) | Residual U-Net · CIFAR10 diffusion | FID: 7.86 -> 12.44 | Hurt |
| add the saved encoder parameters to the decoder in Multi-ResNet (ablation) | Multi-ResNet · Navier-Stokes | r-MSE: 0.0107 -> 0.0040 | Improved |
| add the saved encoder parameters to the decoder in Multi-ResNet (ablation) | Multi-ResNet · Shallow water | r-MSE: 0.4899 -> 0.1493 | Improved |
| add the saved encoder parameters to the decoder in Multi-ResNet (ablation) | Multi-ResNet · WMH segmentation | Dice: 0.8190 -> 0.8346 | Improved |
| add the saved encoder parameters to the decoder in Multi-ResNet (ablation) | Multi-ResNet · CIFAR10 diffusion | FID: 14.87 -> 12.44 | Improved |
| allocate saved parameters via more hidden channels (v2) instead of more residual blocks per resolution (v1) (ablation) | Multi-ResNet · Navier-Stokes and Shallow water | r-MSE | Hurt |
| train with staged multi-resolution training instead of single-stage (ablation) | Residual U-Net diffusion · CIFAR10 | FID: 7.858 -> 8.33 | No clear effect |
Closest papers by meaning
- U-Nets as Belief Propagation: Efficient Classification, Denoising, and Diffusion in Generative Hierarchical Models (ICLR 2025 · similarity 0.89)
- UDT: Reconciling U-Nets and Diffusion Transformers with Data-Adaptive Token Reduction (NeurIPS 2026 · similarity 0.87)
- FreeU: Free Lunch in Diffusion U-Net (CVPR 2024 · similarity 0.86)
- U-DiTs: Downsample Tokens in U-Shaped Diffusion Transformers (NeurIPS 2024 · similarity 0.86)
- U-KAN Makes Strong Backbone for Medical Image Segmentation and Generation (AAAI 2025 · similarity 0.86)
- Neural Collapse is Globally Optimal in Deep Regularized ResNets and Transformers (NeurIPS 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
