Research map / Image generation / Super-resolution
Outlier-Aware Post-Training Quantization for Image Super-Resolution
Hailing Wang, Jianglin Lu, Yitian Zhang, 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 piecewise linear quantizer (PLQ) to MinMax (ablation) | EDSR W4A4 · Set5 | PSNR: 26.83 -> 30.50 | Improved |
| add PLQ to MinMax (ablation) | EDSR W4A4 · Set14 | PSNR: 25.04 -> 27.71 | Improved |
| add PLQ to MinMax (ablation) | EDSR W4A4 · Urban100 | PSNR: 23.12 -> 25.12 | Improved |
| add sensitivity-aware finetuning (SAFT) to MinMax (ablation) | EDSR W4A4 · Set5 | PSNR: 26.83 -> 29.87 | Improved |
| replace vanilla finetuning with SAFT (ablation) | MinMax EDSR W4A4 · Set5 | PSNR: 29.45 -> 29.87 | Improved |
| combine PLQ and SAFT (ablation) | EDSR W4A4 · Set5 | PSNR: 30.50 -> 31.54 | Improved |
| combine PLQ and SAFT instead of SAFT only (ablation) | EDSR W4A4 · Urban100 | PSNR: 24.57 -> 25.61 | Improved |
| use Ours instead of AdaBM | RDN W4A4 · Urban100 | PSNR: 23.38 -> 25.93 | Improved |
| use MinMax instead of Percentile (preserving outliers) | EDSR W4A4 · Set5 | PSNR: 24.03 -> 26.83 | Improved |
| clip 1% of activation outliers in full-precision model | EDSR · visual quality | color fidelity | Hurt |
| use Ours PTQ instead of QAT PAMS | EDSR W4A4 · Set5 | PSNR: 31.59 -> 31.79 | Improved |
| use Ours PTQ instead of QAT ODM | EDSR W4A4 · Set5 | PSNR: 32.00 -> 31.79 | Hurt |
Closest papers by meaning
- Toward Accurate Post-Training Quantization for Image Super Resolution (CVPR 2023 · similarity 0.93)
- 2DQuant: Low-bit Post-Training Quantization for Image Super-Resolution (NeurIPS 2024 · similarity 0.91)
- QuantDemoire: Quantization with Outlier Aware for Image Demoiréing (NeurIPS 2026 · similarity 0.88)
- PassionSR: Post-Training Quantization with Adaptive Scale in One-Step Diffusion based Image Super-Resolution (CVPR 2025 · similarity 0.88)
- AdaBM: On-the-Fly Adaptive Bit Mapping for Image Super-Resolution (CVPR 2024 · similarity 0.88)
- Outlier-aware Slicing for Post-Training Quantization in Vision Transformer (ICML 2024 · similarity 0.88)
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.
