Research map / Image generation / Super-resolution
CADyQ: Content-Aware Dynamic Quantization for Image Super-Resolution
Cheeun Hong, Sungyong Baik, Heewon Kim, Seungjun Nah, Kyoung Mu Lee
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 CADyQ dynamic bit-width instead of fixed 8-bit PAMS | IDN x2 · Urban100 x2 | FQR / PSNR: FQR 8.00 -> 5.22; PSNR 31.39 -> 31.54 | Improved |
| use CADyQ instead of 4-bit DAQ | IDN x2 · Urban100 x2 | PSNR: 29.15 -> 31.54 | Improved |
| use CADyQ instead of 8-bit PAMS | CARN x2 · Urban100 x2 | PSNR: 31.74 -> 31.87 | Improved |
| apply CADyQ on top of the LinQ quantization function | LinQ 8-bit · Urban100 x4 | FQR / PSNR: IDN 8.00 -> 6.85 FQR, PSNR 25.47 -> 25.60 | Improved |
| apply CADyQ on top of fully quantized FQSR | EDSR-baseline FQSR · Urban100 x4 | BitOPs / PSNR: BitOPs 19.8G -> 17.6G; PSNR 25.92 -> 25.90 | Improved |
| add candidate bit-width 6 to {4, 8} (ablation) | CARN-CADyQ · Urban100 x4 | PSNR: 25.90 -> 25.94; FQR 5.02 -> 5.32 | Improved |
| add candidate bit-width 2 to {4, 6, 8} (ablation) | CARN-CADyQ · Urban100 x4 | PSNR: 25.94 -> 25.61; FQR 5.32 -> 3.45 | Hurt |
| use smaller inference patches (192 to 48) (ablation) | CARN-CADyQ · Urban100 x4 | BitOPs / PSNR: BitOPs 3.36G -> 3.12G; PSNR 25.95 -> 25.92 | Improved |
| use CADyQ patch-wise with bit selector instead of PAMS 8-bit image inference | CARN · 4K image | latency: 216.0 ms -> 202.9 ms | Improved |
| increase w_reg (ablation) | CARN-CADyQ · DIV2K validation | PSNR | Hurt |
Later papers that compared with this method
1 later accepted paper reports CADYQ as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Towards Robust Full Low-bit Quantization of Super Resolution Networks (ECCV 2024 · better on 2 of 2 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- QuantSR: Accurate Low-bit Quantization for Efficient Image Super-Resolution (NeurIPS 2023 · similarity 0.86)
- Thinking in Granularity: Dynamic Quantization for Image Super-Resolution by Intriguing Multi-Granularity Clues (AAAI 2025 · similarity 0.85)
- PassionSR: Post-Training Quantization with Adaptive Scale in One-Step Diffusion based Image Super-Resolution (CVPR 2025 · similarity 0.83)
- Text-Aware Real-World Image Super-Resolution via Diffusion Model with Joint Segmentation Decoders (NeurIPS 2025 · similarity 0.82)
- AdaBM: On-the-Fly Adaptive Bit Mapping for Image Super-Resolution (CVPR 2024 · similarity 0.82)
- Towards Realistic Data Generation for Real-World Super-Resolution (ICLR 2025 · similarity 0.82)
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.
