Research map / Image generation / Super-resolution
A Benchmark for Chinese-English Scene Text Image Super-Resolution
Jianqi Ma, Zhetong Liang, Wangmeng Xiang, Xi Yang, Lei Zhang
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 |
|---|---|---|---|
| train on Real-CE instead of TextZoom | RRDB · Real-CE test (4x) | ACC: 0.1463 -> 0.2920 | Improved |
| train on Real-CE instead of RealSR | RRDB · Real-CE test (4x) | ACC: 0.2590 -> 0.2920 | Improved |
| train on Real-CE instead of TextZoom | TBSRN · Real-CE test (4x) | ACC: 0.2246 -> 0.2527 | Improved |
| add pixel-level edge-aware loss to L1 (ablation) | RRDB · Real-CE test (4x) | PSNR: 20.23 -> 20.42 | Improved |
| add pixel-level edge-aware loss to L1 (ablation) | RRDB · Real-CE test (4x) | ACC: 0.2920 -> 0.2914 | No clear effect |
| add feature-level edge-aware loss (ablation) | RRDB · Real-CE test (4x) | ACC: 0.2914 -> 0.3093 | Improved |
| add feature-level edge-aware loss (ablation) | RRDB · Real-CE test (4x) | LPIPS: 0.2630 -> 0.2031 | Improved |
| add feature-level edge-aware loss (ablation) | RRDB · Real-CE test (4x) | PSNR: 20.42 -> 20.14 | Hurt |
| add both edge-aware losses (ablation) | TATT · Real-CE test (4x) | NED: 0.4342 -> 0.4549 | Improved |
Closest papers by meaning
- Restore Text First, Enhance Image Later: Two-Stage Scene Text Image Super-Resolution with Glyph Structure Guidance (CVPR 2026 · similarity 0.90)
- Enhanced Generative Structure Prior for Chinese Text Image Super-Resolution (TPAMI 2025 · similarity 0.89)
- Text-Aware Real-World Image Super-Resolution via Diffusion Model with Joint Segmentation Decoders (NeurIPS 2025 · similarity 0.87)
- Beyond Cropped Regions: New Benchmark and Corresponding Baseline for Chinese Scene Text Retrieval in Diverse Layouts (ICML 2025 · similarity 0.87)
- GlyphSR: A Simple Glyph-Aware Framework for Scene Text Image Super-Resolution (AAAI 2025 · similarity 0.86)
- OTE: Exploring Accurate Scene Text Recognition Using One Token (CVPR 2024 · 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.
