Research map / Image generation / Super-resolution
DCDM: Diffusion-Conditioned-Diffusion Model for Scene Text Image Super-Resolution
Shrey Singh, Prateek Keserwani, Masakazu Iwamura, Partha Pratim Roy
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 latent text diffusion model as conditioning for the image diffusion model (ablation) | DCDM · TextZoom (CRNN, easy) | recognition accuracy: 65.2% -> 65.7% | Improved |
| add latent text diffusion model as conditioning (ablation) | DCDM · TextZoom (MORAN, medium) | recognition accuracy: 61.7% -> 63.5% | Improved |
| add latent text diffusion model as conditioning (ablation) | DCDM · TextZoom (ASTER, easy) | recognition accuracy: 80.4% -> 81.8% | Improved |
| add latent text diffusion model as conditioning (ablation) | DCDM · TextZoom (CRNN, hard) | recognition accuracy: 41.3% -> 41.4% | Improved |
| use DCDM without ground-truth text in training instead of TCDM which uses text ground truth | scene text SR · TextZoom (CRNN average) | recognition accuracy: 55.7% -> 55.5% | No clear effect |
| use DCDM instead of TCDM on ASTER hard subset | scene text SR · TextZoom (ASTER, hard) | recognition accuracy: 50.1% -> 47.4% | Hurt |
| use DCDM instead of text-prior baselines (TATT, C3-STISR) on TextZoom | scene text SR · TextZoom (CRNN average) | recognition accuracy: 53.7% -> 55.5% (vs C3-STISR) | Improved |
| use DCDM instead of prior STISR methods on Real-CE | scene text SR · Real-CE x4 / x2 | PSNR, SSIM, ACC: x4 trained on Real-CE ACC 25.27 -> 25.49 (vs TBSRN) | Improved |
Closest papers by meaning
- Diffusion-based Blind Text Image Super-Resolution (CVPR 2024 · similarity 0.90)
- Text-Aware Real-World Image Super-Resolution via Diffusion Model with Joint Segmentation Decoders (NeurIPS 2025 · similarity 0.89)
- Layout-Agnostic Scene Text Image Synthesis with Diffusion Models (CVPR 2024 · similarity 0.89)
- PRISM: Prior Rectification and Uncertainty-Aware Structure Modeling for Diffusion-Based Text Image Super-Resolution (NeurIPS 2026 · similarity 0.88)
- DreamText: High Fidelity Scene Text Synthesis (CVPR 2025 · similarity 0.88)
- Get What You Want, Not What You Don't: Image Content Suppression for Text-to-Image Diffusion Models (ICLR 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.
