Research map / Image generation / Super-resolution
ClassSR: A General Framework to Accelerate Super-Resolution Networks by Data Characteristic
Check 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 |
|---|---|---|---|
| route sub-images through difficulty-specific branches | FSRCNN · Test8K | FLOPs: 468M(100%) -> 238M(51%) | Helped |
| route sub-images through difficulty-specific branches | CARN · Test8K | FLOPs: 1.15G(100%) -> 608M(53%) | Helped |
| route sub-images through difficulty-specific branches | SRResNet · Test8K | FLOPs: 5.20G(100%) -> 2.70G(52%) | Helped |
| route sub-images through difficulty-specific branches | RCAN · Test8K | FLOPs: 32.60G(100%) -> 16.36G(50%) | Helped |
| route sub-images through difficulty-specific branches | RCAN · Test8K | PSNR: 33.76dB -> 33.73dB | Hurt |
| remove Class-Loss (ablation) | ClassSR-FSRCNN · DIV2K validation during training | PSNR | Hurt |
| add Class-Loss (ablation) | ClassSR-FSRCNN · DIV2K validation during training | classification stability | Helped |
| remove Average-Loss (ablation) | ClassSR-FSRCNN · DIV2K validation during training | FLOPs | Hurt |
| add Average-Loss (ablation) | ClassSR-FSRCNN · DIV2K validation during training | branch utilization | Helped |
| increase the number of branches from 2 to 5 (ablation) | ClassSR-FSRCNN · Test8K | PSNR: 32.72dB -> 32.74dB | Helped |
| vary branch channel and layer configurations (ablation) | ClassSR-SRResNet · Test2K | PSNR: 26.20dB; 26.20dB | No clear effect |
| replace SR-Module with DnCNN branches under ClassSR | DnCNN · Test2K | PSNR: 31.20dB -> 31.23dB | Helped |
Closest papers by meaning
- FSR: A General Frequency-Oriented Framework to Accelerate Image Super-resolution Networks (AAAI 2023 · similarity 0.90)
- Accurate Image Super-Resolution Using Very Deep Convolutional Networks (CVPR 2016 · similarity 0.90)
- Image Processing GNN: Breaking Rigidity in Super-Resolution (CVPR 2024 · similarity 0.87)
- SeD: Semantic-Aware Discriminator for Image Super-Resolution (CVPR 2024 · similarity 0.87)
- Low-Res Leads the Way: Improving Generalization for Super-Resolution by Self-Supervised Learning (CVPR 2024 · similarity 0.87)
- Learning Coupled Dictionaries from Unpaired Data for Image Super-Resolution (CVPR 2024 · similarity 0.86)
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.