Research map / Image generation / Image restoration
Maximum a posteriori natural scene reconstruction from retinal ganglion cells with deep denoiser priors
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 |
|---|---|---|---|
| replace 1/f Gaussian prior with deep denoiser prior (ablation) | MAP-GLM reconstruction · Retina 1 test | MS-SSIM: 0.557 -> 0.689 | Helped |
| replace 1/f Gaussian prior with deep denoiser prior (ablation) | MAP-GLM reconstruction · Retina 2 test | PSNR: 17.3 -> 18.5 | Helped |
| replace LNP encoding model with GLM (ablation) | MAP-dCNN reconstruction · Retina 1 test | MS-SSIM: 0.635 -> 0.689 | Helped |
| replace LNP encoding model with GLM (ablation) | MAP-dCNN reconstruction · Retina 2 test | PSNR: 17.0 -> 18.5 | Helped |
| MAP-GLM-dCNN vs L-CAE regression baseline | RGC image reconstruction · Retina 1 test | MS-SSIM: 0.665 -> 0.689 | Helped |
| MAP-GLM-dCNN vs L-CAE regression baseline | RGC image reconstruction · Retina 1 test | PSNR: 19.9 -> 19.5 | Hurt |
| MAP-GLM-dCNN vs Kim et al. regression method | RGC image reconstruction · Retina 2 test | MS-SSIM: 0.675 -> 0.668 | No clear effect |
| MAP-GLM-dCNN vs Kim et al. regression method | RGC image reconstruction · Retina 1 test | PSNR: 20.2 -> 19.5 | Hurt |
Closest papers by meaning
- Deep Learning Models of the Retinal Response to Natural Scenes (NeurIPS 2016 · similarity 0.91)
- Neural Networks for Efficient Bayesian Decoding of Natural Images from Retinal Neurons (NeurIPS 2017 · similarity 0.91)
- Natural image synthesis for the retina with variational information bottleneck representation (NeurIPS 2022 · similarity 0.91)
- SpikeReveal: Unlocking Temporal Sequences from Real Blurry Inputs with Spike Streams (NeurIPS 2024 · similarity 0.86)
- Finding Visual Saliency in Continuous Spike Stream (AAAI 2024 · similarity 0.86)
- MonkeySee: Space-time-resolved reconstructions of natural images from macaque multi-unit activity (NeurIPS 2024 · 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.