Research map / Image generation / Super-resolution
MuLUT: Cooperating Multiple Look-Up Tables for Efficient Image Super-Resolution
Jiacheng Li, Chang Chen, Zhen Cheng, Zhiwei Xiong
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 complementary indexing patterns D and Y (5x5 RF) to a single LUT (ablation) | SR-LUT-S · Manga109 4x SR | PSNR: 26.80 -> 27.52 | Improved |
| add MuLUT-Y to MuLUT-SD (ablation) | MuLUT-SD · Manga109 4x SR | PSNR: 27.38 -> 27.52 | Improved |
| add MuLUT-D to SR-LUT-S (ablation) | SR-LUT-S · Manga109 4x SR | PSNR: 26.80 -> 27.38 | Improved |
| add LUT re-indexing to cascaded 2-stage LUTs (ablation) | MuLUT-S-X2 · Manga109 4x SR | PSNR: 27.02 -> 27.39 | Improved |
| cascade more stages (X2 to X4) (ablation) | MuLUT-S · Manga109 4x SR | PSNR: 27.39 -> 27.66 | Improved |
| combine complementary and hierarchical indexing (SDY-X2 vs SDY) (ablation) | MuLUT · Manga109 4x SR | PSNR: 27.52 -> 27.90 | Improved |
| add LUT-aware finetuning to 4bit LUT (ablation) | SR-LUT-S · Manga109 4x SR | PSNR: 26.80 -> 26.94 | Improved |
| add LUT-aware finetuning to 3bit LUT (ablation) | SR-LUT-S · Manga109 4x SR | PSNR: 26.76 -> 26.85 | Improved |
| add LUT-aware finetuning (ablation) | MuLUT-SDY-X2 · Manga109 4x SR | PSNR: 27.83 -> 27.90 | Improved |
| use MuLUT-SDY-X2 instead of SR-LUT-S | SR-LUT-S · Manga109 4x SR | PSNR: 26.80 -> 27.90 | Improved |
| use MuLUT-SDY-X2 instead of SR-LUT-S | SR-LUT-S · 1280x720 2x SR energy | energy cost: 74.2M -> 278.5M | Hurt |
| use MuLUT-SDY-X2 instead of FSRCNN | FSRCNN · 1280x720 2x SR energy | energy cost: 28.1G -> 278.5M | Improved |
Closest papers by meaning
- Efficient Look-Up Table from Expanded Convolutional Network for Accelerating Image Super-resolution (AAAI 2024 · similarity 0.89)
- Edge-SD-SR: Low Latency and Parameter Efficient On-device Super-Resolution with Stable Diffusion via Bidirectional Conditioning (CVPR 2025 · similarity 0.86)
- TinyLUT: Tiny Look-Up Table for Efficient Image Restoration at the Edge (NeurIPS 2024 · similarity 0.85)
- AutoLUT: LUT-Based Image Super-Resolution with Automatic Sampling and Adaptive Residual Learning (CVPR 2025 · similarity 0.85)
- PocketSR: The Super-Resolution Expert in Your Pocket Mobiles (NeurIPS 2025 · similarity 0.84)
- CFAT: Unleashing Triangular Windows for Image Super-resolution (CVPR 2024 · similarity 0.84)
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.
