Research map / Image generation / Super-resolution
StyleSRN: Scene Text Image Super-Resolution with Text Style Embedding
Shengrong Yuan, Runmin Wang, Ke Hao, Xuqi Ma, Changxin Gao, Li Liu, Nong Sang
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 text style loss (ablation) | StyleSRN · TextZoom avg (CRNN) | accuracy (%): 54.9 -> 57.4 | Improved |
| add text style loss (ablation) | TSRN · TextZoom avg (CRNN) | accuracy (%): 41.4 -> 43.7 | Improved |
| add text style loss (ablation) | TATT · TextZoom avg (CRNN) | accuracy (%): 52.6 -> 53.8 | Improved |
| add text style loss (ablation) | TSRN · TextZoom avg | PSNR: 21.42 -> 21.56 | Improved |
| add SCFB (ablation) | TSRN · TextZoom avg (CRNN) | accuracy (%): 41.4 -> 43.5 | Improved |
| add SEB on top of SCFB (ablation) | TSRN · TextZoom avg (CRNN) | accuracy (%): 43.5 -> 44.4 | Improved |
| add SEB on top of SCFB (ablation) | TATT · TextZoom avg | SSIM: 0.7790 -> 0.7785 | Hurt |
| add SCFB and SEB (ablation) | LEMMA · TextZoom avg (CRNN) | accuracy (%): 56.3 -> 57.0 | Improved |
| use CRNN instead of VGG for style loss features (ablation) | StyleSRN · TextZoom avg (CRNN) | accuracy (%): 56.3 -> 57.4 | Improved |
| use 3 SEBs instead of none (ablation) | StyleSRN · TextZoom avg | PSNR: 20.76 -> 21.82 | Improved |
| use 5 SEBs instead of 3 (ablation) | StyleSRN · TextZoom avg (CRNN) | accuracy (%): 57.4 -> 55.9 | Hurt |
| use 5 SCFBs instead of none (ablation) | StyleSRN · TextZoom avg | PSNR: 20.67 -> 21.82 | Improved |
Closest papers by meaning
- Text-Aware Real-World Image Super-Resolution via Diffusion Model with Joint Segmentation Decoders (NeurIPS 2025 · similarity 0.87)
- GlyphSR: A Simple Glyph-Aware Framework for Scene Text Image Super-Resolution (AAAI 2025 · similarity 0.84)
- Boosting Semi-Supervised Scene Text Recognition via Viewing and Summarizing (NeurIPS 2024 · similarity 0.84)
- SRSR: Enhancing Semantic Accuracy in Real-World Image Super-Resolution with Spatially Re-Focused Text-Conditioning (NeurIPS 2025 · similarity 0.84)
- OTE: Exploring Accurate Scene Text Recognition Using One Token (CVPR 2024 · similarity 0.84)
- PRISM: Prior Rectification and Uncertainty-Aware Structure Modeling for Diffusion-Based Text Image Super-Resolution (NeurIPS 2026 · 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.
