Research map / 3D vision / Face and shape
Residual Multiplicative Filter Networks for Multiscale Reconstruction
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Add residual connections to staged Bacon (ablation) | Bacon · Natural images, scale 3 | Mean absolute Fourier spectrum difference: 17.97 $\pm$ 4.27 -> 9.23 $\pm$ 2.73 | Helped |
| Add the proposed frequency initialization to residual Bacon (ablation) | Residual MFN · Natural images, scale 3 | Mean absolute Fourier spectrum difference: 9.23 $\pm$ 2.73 -> 4.32 $\pm$ 1.35 | Helped |
| Add residual connections and proposed frequency initialization | Bacon · Natural images, image generalization | PSNR (dB): 28.46 $\pm$ 3.61 -> 28.33 $\pm$ 4.42 | No clear effect |
| Add residual connections to staged Bacon (ablation) | Bacon · Text images, scale 3 | Mean absolute Fourier spectrum difference: 13.79 $\pm$ 4.60 -> 8.11 $\pm$ 2.60 | Helped |
| Add the proposed frequency initialization to residual Bacon (ablation) | Residual MFN · Text images, scale 3 | Mean absolute Fourier spectrum difference: 8.11 $\pm$ 2.60 -> 3.56 $\pm$ 1.16 | Helped |
| Add residual connections and proposed frequency initialization | Bacon · Text images, image generalization | PSNR (dB): 31.57 $\pm$ 2.29 -> 32.04 $\pm$ 2.34 | Helped |
| Train with coarse-to-fine stages rather than supervising only the full-scale reconstruction (ablation) | Residual MFN cryo-EM reconstruction · Synthetic cryo-EM reconstruction | Structure reconstruction / pose optimization | Helped |
| Use λ₂ = 2 and λ₁ = 0.3 rather than the other investigated λ settings (ablation) | Residual MFN · Image fitting and cryo-EM reconstruction | Test loss / FSC | Helped |
Closest papers by meaning
- DEFT: Efficient Fine-tuning of Diffusion Models by Learning the Generalised $h$-transform (NeurIPS 2024 · similarity 0.85)
- Boosting ViT-based MRI Reconstruction from the Perspectives of Frequency Modulation, Spatial Purification, and Scale Diversification (AAAI 2025 · similarity 0.83)
- A Variational Perspective on Solving Inverse Problems with Diffusion Models (ICLR 2024 · similarity 0.83)
- DRACO: A Denoising-Reconstruction Autoencoder for Cryo-EM (NeurIPS 2024 · similarity 0.83)
- Multi-Cross Sampling and Frequency-Division Reconstruction for Image Compressed Sensing (AAAI 2024 · similarity 0.83)
- Fast samplers for Inverse Problems in Iterative Refinement models (NeurIPS 2024 · similarity 0.83)
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.