Research map / Image generation / Super-resolution
LatticeNet: Towards Lightweight Image Super-resolution with Lattice Block
Xiaotong Luo, Yuan Xie, Yulun Zhang, Yanyun Qu, Cuihua Li, Yun Fu
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 lattice block to baseline (ablation) | LatticeNet baseline · Set5 2x SR | PSNR: 37.84 -> 37.96 | Improved |
| add lattice block to baseline (ablation) | LatticeNet baseline · Set14 2x SR | PSNR: 33.43 -> 33.59 | Improved |
| add backward fusion module to baseline (ablation) | LatticeNet baseline · Set5 2x SR | PSNR: 37.84 -> 37.88 | Improved |
| add backward fusion module to baseline (ablation) | LatticeNet baseline · Set14 2x SR | PSNR: 33.43 -> 33.47 | Improved |
| use BFM instead of 1x1 conv fusion (ablation) | LatticeNet baseline · Set5 2x SR | PSNR: 37.85 -> 37.88 | Improved |
| combine LB and BFM (ablation) | LatticeNet baseline · Set5 2x SR | PSNR: 37.84 -> 38.01 | Improved |
| combine mean pooling and std pooling for combination coefficients (vs MP only) (ablation) | LatticeNet · Set5 2x SR | PSNR: 37.985 -> 38.012 | Improved |
| combine mean pooling and std pooling (vs SDP only) (ablation) | LatticeNet · Urban100 2x SR | PSNR: 32.099 -> 32.145 | Improved |
| replace four residual blocks with lattice blocks in SRResNet (ablation) | SRResNet · Set5 4x SR | PSNR: 32.18 -> 32.20 | No clear effect |
| replace residual blocks with lattice blocks in RCAN (ablation) | RCAN · Set5 4x SR | PSNR: 32.63 -> 32.59 | No clear effect |
| increase number of lattice blocks from 2 to 10 (ablation) | LatticeNet · Set5 4x SR | PSNR: 32.03 -> 32.35 | Improved |
| increase number of lattice blocks from 4 to 8 (ablation) | LatticeNet · Set5 4x SR | PSNR: 32.18 -> 32.29 | Improved |
Closest papers by meaning
- Lattice Network for Lightweight Image Restoration (TPAMI 2022 · similarity 0.92)
- Fast, Accurate, and Lightweight Super-Resolution with Cascading Residual Network (ECCV 2018 · similarity 0.90)
- Multi-scale Residual Network for Image Super-Resolution (ECCV 2018 · similarity 0.89)
- Efficient Residual Dense Block Search for Image Super-Resolution (AAAI 2020 · similarity 0.89)
- Learning Efficient Image Super-Resolution Networks via Structure-Regularized Pruning (ICLR 2022 · similarity 0.88)
- Lightweight Image Super-Resolution via Flexible Meta Pruning (ICML 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.
