Research map / Image generation / Text-to-image diffusion
One-Shot Diffusion Mimicker for Handwritten Text Generation
Gang Dai, Yifan Zhang, Quhui Ke, Qiangya Guo, Shuangping Huang
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 the Laplacian high-frequency style branch (ablation) | One-DM · IAM OOV-U | FID: 108.44 -> 104.52 | Improved |
| add the gate mechanism to suppress background noise (ablation) | One-DM · IAM OOV-U | FID: 108.44 -> 105.73 | Improved |
| combine Laplacian branch and gate (ablation) | One-DM · IAM OOV-U | FID: 108.44 -> 102.75 | Improved |
| use LapNCE contrastive loss together with high-frequency images rather than either alone (ablation) | One-DM · IAM OOV-U | FID: 106.16 / 107.58 -> 104.52 | Improved |
| use cross-attention style-content fusion instead of concatenation or timestep injection (ablation) | One-DM · IAM OOV-U | FID: 104.81 / 104.26 -> 102.75 | Improved |
| use a single reference sample instead of 15 for style | One-DM vs few-shot VATr · IAM OOV-U | FID: 108.76 -> 102.75 | Improved |
| use style reference words of length one letter (ablation) | One-DM · IAM OOV-U | FID | Hurt |
| use style references longer than four letters (ablation) | One-DM · IAM OOV-U | FID | Improved |
| train OCR on One-DM synthetic data instead of VATr synthetic data | OCR recognizer · IAM test | CER (%): 26.98 -> 11.69 | Improved |
| Laplacian filter instead of other high-frequency filters (ablation) | One-DM · IAM OOV-U | FID | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- VATr — VATr++: Choose Your Words Wisely for Handwritten Text Generation (TPAMI 2024) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- FontDiffuser: One-Shot Font Generation via Denoising Diffusion with Multi-Scale Content Aggregation and Style Contrastive Learning (AAAI 2024 · similarity 0.86)
- Decoupling Layout from Glyph in Online Chinese Handwriting Generation (ICLR 2025 · similarity 0.86)
- Zero-Shot Styled Text Image Generation, but Make It Autoregressive (CVPR 2025 · similarity 0.85)
- Generate Like Experts: Multi-Stage Font Generation by Incorporating Font Transfer Process into Diffusion Models (CVPR 2024 · similarity 0.85)
- InstaFlow: One Step is Enough for High-Quality Diffusion-Based Text-to-Image Generation (ICLR 2024 · similarity 0.84)
- Layout-Agnostic Scene Text Image Synthesis with Diffusion Models (CVPR 2024 · 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.
