Research map / Vision-language and multimodal / Human and embodied modeling
Superpixel-informed Implicit Neural Representation for Multi-Dimensional Data
Jia-Yi Li, Xi-Le Zhao, Jian-Li Wang, Chao Wang, Min Wang
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 |
|---|---|---|---|
| use generalized superpixels instead of pixels as INR basic units (ablation) | S-INR · Lehavim denoising (Case2) | PSNR: 29.348 -> 31.475 | Improved |
| add shared dictionary matrix to superpixel INR (ablation) | S-INR · Lehavim denoising (Case2) | PSNR: 31.475 -> 31.997 | Improved |
| add exclusive attention-based MLPs (with dictionary) (ablation) | S-INR · Lehavim denoising (Case2) | PSNR: 31.997 -> 32.394 | Improved |
| add exclusive attention MLPs without the shared dictionary (ablation) | S-INR · Lehavim denoising (Case2) | PSNR: 31.475 -> 31.012 | Hurt |
| add shared dictionary to full S-INR (ablation) | S-INR · Lehavim denoising (Case2) | PSNR: 31.012 -> 32.394 | Improved |
| S-INR vs SIREN on image reconstruction | S-INR · Kodim image reconstruction | PSNR: 33.052 -> 36.077 | Improved |
| S-INR vs SIREN on 3D surface completion | S-INR · 3D surface completion Scene1 | NRMSE: 0.109 -> 0.074 | Improved |
| S-INR vs best baseline on image completion | S-INR · Mor MSI completion (Case2, rate 0.05) | SSIM: 0.897 -> 0.900 | No clear effect |
Closest papers by meaning
- Implicit Representations via Operator Learning (ICML 2024 · similarity 0.87)
- Grids Often Outperform Implicit Neural Representation at Compressing Dense Signals (NeurIPS 2025 · similarity 0.86)
- VI^3NR: Variance Informed Initialization for Implicit Neural Representations (CVPR 2025 · similarity 0.85)
- SINR: Sparsity Driven Compressed Implicit Neural Representations (CVPR 2025 · similarity 0.85)
- DVI:A Derivative-based Vision Network for INR (ICML 2025 · similarity 0.85)
- HIIF: Hierarchical Encoding based Implicit Image Function for Continuous Super-resolution (CVPR 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.
