Research map / Image generation / Text-to-image diffusion
TextMaster: A Unified Framework for Realistic Text Editing via Glyph-Style Dual-Control
Zhenyu Yan, Jian Wang, Aoqiang Wang, Yuhan Li, Wenxiang Shang, Zhu Hangcheng
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 token-separated text encoder (ablation) | TextMaster (cumulative) · Ablation (500 images) | OCR accuracy: 0.5098 -> 0.6832 | Improved |
| add token-separated text encoder (ablation) | TextMaster (cumulative) · Ablation (500 images) | FID: 16.26 -> 14.58 | Improved |
| add perceptual loss (ablation) | TextMaster (cumulative) · Ablation (500 images) | OCR accuracy: 0.6832 -> 0.7451 | Improved |
| add perceptual loss (ablation) | TextMaster (cumulative) · Ablation (500 images) | FID: 14.58 -> 12.95 | Improved |
| add attention IoU loss (ablation) | TextMaster (cumulative) · Ablation (500 images) | NED: 0.7638 -> 0.8495 | Improved |
| add attention IoU loss (ablation) | TextMaster (cumulative) · Ablation (500 images) | FID: 12.95 -> 13.02 | No clear effect |
| add in-context reference (ablation) | TextMaster (cumulative) · Ablation (500 images) | OCR accuracy: 0.7816 -> 0.8162 | Improved |
| add in-context reference (ablation) | TextMaster (cumulative) · Ablation (500 images) | FID: 13.02 -> 11.65 | Improved |
| add style adapter (ablation) | TextMaster (cumulative) · Ablation (500 images) | OCR accuracy: 0.8162 -> 0.8860 | Improved |
| add style adapter (ablation) | TextMaster (cumulative) · Ablation (500 images) | LPIPS: 0.0338 -> 0.0286 | Improved |
| use TextMaster instead of TextCtrl | TextMaster · AnyText-Benchmark English | FID: 6.01 -> 10.35 | Hurt |
| use TextMaster instead of AnyText2 | TextMaster · AnyText-Benchmark Chinese | OCR accuracy: 0.6133 -> 0.9257 | 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.
- TextCtrl — TextCtrl: Diffusion-based Scene Text Editing with Prior Guidance Control (NeurIPS 2024) · better on 4 of 5 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
- GlyphMastero: A Glyph Encoder for High-Fidelity Scene Text Editing (CVPR 2025 · similarity 0.89)
- TextCtrl: Diffusion-based Scene Text Editing with Prior Guidance Control (NeurIPS 2024 · similarity 0.89)
- GlyphAnchor: Enhancing Visual Text Rendering via Position-Anchored Glyph Priors (NeurIPS 2026 · similarity 0.88)
- SteerVTE: Seamless Video Text Editing with Style and Glyph Control (NeurIPS 2026 · similarity 0.88)
- Text-Driven Image Editing via Learnable Regions (CVPR 2024 · similarity 0.86)
- FeedEdit: Text-Based Image Editing with Dynamic Feedback Regulation (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.
