Research map / Architectures and efficiency / Quantization
CABM: Content-Aware Bit Mapping for Single Image Super-Resolution Network With Large Input
Senmao Tian, Ming Lu, Jiaming Liu, Yandong Guo, Yurong Chen, Shunli Zhang
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 |
|---|---|---|---|
| replace CADyQ's MLP bit selector with CABM edge-to-bit lookup tables | CARN · Urban100, x4 | BitOPs: 3.13G -> 1.57G | Improved |
| replace CADyQ's MLP bit selector with CABM edge-to-bit lookup tables | EDSR · Urban100, x4 | PSNR (dB): 25.90 -> 25.95 | No clear effect |
| replace CADyQ's MLP bit selector with CABM edge-to-bit lookup tables | EDSR · Test2K, x4 | PSNR (dB): 27.54 -> 27.57 | No clear effect |
| replace CADyQ's MLP bit selector with CABM edge-to-bit lookup tables | EDSR · Test4K, x4 | BitOPs: 82.37G -> 47.70G | Improved |
| add fine-tuning after replacing MLP-selected bits with CABM lookup tables (ablation) | EDSR · Urban100, x4 | PSNR (dB): 25.77 -> 25.95 | Improved |
| add fine-tuning after replacing MLP-selected bits with CABM lookup tables (ablation) | IDN · Urban100, x4 | PSNR (dB): 24.32 -> 25.58 | Improved |
| increase the calibration expansion range Δe from 10 to 80 (ablation) | EDSR CABM · Urban100, x4 | PSNR (dB): 25.95 -> 25.87 | Hurt |
| choose maximum- rather than minimum-BitOPs configurations per interval (ablation) | EDSR CABM · Urban100, x4 | PSNR (dB): 25.95 -> 26.04 | Improved |
| choose maximum- rather than minimum-BitOPs configurations per interval (ablation) | EDSR CABM · Urban100, x4 | FAB: 5.90 -> 6.64 | Hurt |
| randomly choose rather than select minimum-BitOPs configurations per interval (ablation) | EDSR CABM · Urban100, x4 | PSNR (dB): 25.95 -> 26.01 | Improved |
| sample subnet difficulty in proportion to BitOPs rather than uniformly (ablation) | EDSR supernet · Urban100 | PSNR (dB): 25.94 -> 26.01 | Improved |
Closest papers by meaning
- AdaBM: On-the-Fly Adaptive Bit Mapping for Image Super-Resolution (CVPR 2024 · similarity 0.88)
- QuantSR+: Pushing the Limit of Quantized Image Super-Resolution Networks (TPAMI 2026 · similarity 0.88)
- QuantSR: Accurate Low-bit Quantization for Efficient Image Super-Resolution (NeurIPS 2023 · similarity 0.87)
- IAFMNet: Information-Aware Feature Modulation for Efficient Super-Resolution (CVPR 2026 · similarity 0.87)
- Edge-SD-SR: Low Latency and Parameter Efficient On-device Super-Resolution with Stable Diffusion via Bidirectional Conditioning (CVPR 2025 · similarity 0.86)
- Thinking in Granularity: Dynamic Quantization for Image Super-Resolution by Intriguing Multi-Granularity Clues (AAAI 2025 · similarity 0.85)
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.
