Research map / Image generation / Text-to-image diffusion
Conditional Text Image Generation With Diffusion Models
Yuanzhi Zhu, Zhaohai Li, Tianwei Wang, Mengchao He, Cong Yao
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 CTIG-DM generated data to CRNN training | CRNN · IAM | WER: 18.15 -> 13.37 | Improved |
| Add random geometric augmentation to CTIG-DM generated data | CTIG-DM with CRNN · IAM | WER: 13.37 -> 12.01 | Improved |
| Add CTIG-DM generated data to CRNN training | CRNN · RIMES | WER: 13.44 -> 8.24 | Improved |
| Train with CTIG-DM generated data | CRNN · STR test datasets average | accuracy: 77.2 -> 78.9 | Improved |
| Train with CTIG-DM generated data | ABINet · STR test datasets average | accuracy: 92.7 -> 93.2 | Improved |
| Add CTIG-DM generated data to recognizer training | CRNN · Real-L STR test datasets average | accuracy: 75.0 -> 79.9 | Improved |
| Add CTIG-DM generated data to recognizer training | ABINet · Real-L STR test datasets average | accuracy: 92.8 -> 94.3 | Improved |
| Add image condition to an unconditional diffusion model (ablation) | CTIG-DM · IAM | FID: 33.42 -> 13.36 | Improved |
| Add text condition to an unconditional diffusion model (ablation) | CTIG-DM · IAM | FID: 33.42 -> 10.92 | Improved |
| Add text condition alongside image condition (ablation) | CTIG-DM · IAM | FID: 13.36 -> 9.76 | Improved |
| Add style condition alongside image and text conditions (ablation) | CTIG-DM · IAM | FID: 9.76 -> 9.34 | Improved |
| Use CTIG-DM rather than SLOGAN for OOV image generation | IAM OOV generation · IAM OOV | FID: 97.81 -> 25.52 | Improved |
Closest papers by meaning
- Layout-Agnostic Scene Text Image Synthesis with Diffusion Models (CVPR 2024 · similarity 0.88)
- Referee Can Play: An Alternative Approach to Conditional Generation via Model Inversion (ICML 2024 · similarity 0.88)
- Brush Your Text: Synthesize Any Scene Text on Images via Diffusion Model (AAAI 2024 · similarity 0.87)
- Diffusion-based Blind Text Image Super-Resolution (CVPR 2024 · similarity 0.87)
- Towards Understanding the Working Mechanism of Text-to-Image Diffusion Model (NeurIPS 2024 · similarity 0.87)
- Exploring the Deep Fusion of Large Language Models and Diffusion Transformers for Text-to-Image Synthesis (CVPR 2025 · similarity 0.86)
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.
