Research map / Image generation / Super-resolution
Fast and Accurate Image Super-Resolution with Deep Laplacian Pyramid Networks
Wei‐Sheng Lai, Jia‐Bin Huang, Narendra Ahuja, Ming–Hsuan Yang
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 |
|---|---|---|---|
| add pyramid structure (ablation) | LapSRN · Set5 4x | PSNR: 30.58 -> 31.28 | Improved |
| add global residual learning (ablation) | LapSRN · Set5 4x | PSNR: 31.10 -> 31.28 | Improved |
| use Charbonnier loss instead of L2 (ablation) | LapSRN · Set5 4x | PSNR: 30.93 -> 31.28 | Improved |
| add pyramid structure (ablation) | LapSRN · Set14 4x | PSNR: 27.61 -> 28.04 | Improved |
| share weights across pyramid levels (half the parameters) (ablation) | LapSRN · Urban100 4x | PSNR: 25.21 -> 25.20 | No clear effect |
| reduce to D2R5 with 112k parameters (ablation) | LapSRN · Urban100 4x | PSNR: 25.21 -> 25.10 | Hurt |
| use shared-source local skip connections (ablation) | LapSRN D5R8 · Urban100 4x | PSNR: 25.26 -> 25.38 | Improved |
| increase recursive blocks R from 2 to 8 with same parameters (ablation) | LapSRN_SS D5 · Urban100 4x | PSNR: 25.23 -> 25.38 | Improved |
| increase depth with D2 (R5 to R20) (ablation) | LapSRN D2 · BSDS100 4x | PSNR: 27.33 -> 27.37 | Improved |
| add multi-scale training (2x,4x,8x) vs 4x only (ablation) | MS-LapSRN · Urban100 4x | PSNR: 25.38 -> 25.51 | Improved |
| add multi-scale training (2x,4x,8x) vs 8x only (ablation) | MS-LapSRN · Urban100 8x | PSNR: 21.95 -> 22.06 | Improved |
| replace distinct-pyramid LAPGAN generator by LapSRN (ablation) | LapSRN · BSDS100 4x | PSNR: 27.09 -> 27.22 | Improved |
Closest papers by meaning
- Deep Laplacian Pyramid Networks for Fast and Accurate Super-Resolution (CVPR 2017 · similarity 0.97)
- Densely Residual Laplacian Super-Resolution (TPAMI 2020 · similarity 0.92)
- Image Super-Resolution Using Deep Convolutional Networks (TPAMI 2016 · similarity 0.91)
- Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network (CVPR 2016 · similarity 0.90)
- Fast, Accurate, and Lightweight Super-Resolution with Cascading Residual Network (ECCV 2018 · similarity 0.90)
- LaMAR: Laplacian Pyramid for Multimodal Adaptive Super Resolution (Student Abstract) (AAAI 2024 · similarity 0.87)
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.
