Research map / Image generation / Text-to-image diffusion
InnoText: A Unified Model for Visual Text Generation and Editing
Haowei Liu, Runze He, Jian Lu, Ao Ma, Run Ling, Ke Cao, Jiasong Feng, Wei Feng, Shuo Lu, Yexing Xu, WANG Yun, Jing Wang and 1 more
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 Font Size-Aware Modulation (FSAM) (ablation) | InnoText · InnoText-Bench generation | LPIPS: 0.5928 -> 0.5311 | Improved |
| add FSAM (ablation) | InnoText · InnoText-Bench editing | Sen. Acc: 0.5672 -> 0.6374 | Improved |
| add Small-Character Aware Augmentation (ablation) | InnoText · InnoText-Bench editing | Sen. Acc: 0.5963 -> 0.6374 | Improved |
| add Task-Specific Region Weighted Loss (ablation) | InnoText · InnoText-Bench editing | Sen. Acc: 0.6180 -> 0.6374 | Improved |
| add Task-Specific Region Weighted Loss (ablation) | InnoText · InnoText-Bench editing | NED: 0.7533 -> 0.7528 | No clear effect |
| fine-tune AnyText2 on InnoText-30K (ablation) | AnyText2 · InnoText-Bench | Sen. Acc: edit 0.5882, gen 0.5358 vs InnoText 0.6374 / 0.5907 | Mixed |
| train on AnyWord-3M instead of InnoText-30K | InnoText · AnyText-Benchmark editing (English) | Sen. Acc: 0.7988 -> 0.8224 | Improved |
| increase LoRA rank from 32 to 128 (ablation) | InnoText · InnoText-Bench | Sen. Acc: 0.5902 -> 0.6586 | Improved |
| increase LoRA rank from 128 to 256 (ablation) | InnoText · InnoText-Bench generation | Sen. Acc: 0.5907 -> 0.5889 | No clear effect |
| unified model vs TextFlux on generation | InnoText · InnoText-Bench generation | Sen. Acc: 0.0071 -> 0.6586 | Improved |
Closest papers by meaning
- AnyText: Multilingual Visual Text Generation and Editing (ICLR 2024 · similarity 0.89)
- DreamText: High Fidelity Scene Text Synthesis (CVPR 2025 · similarity 0.89)
- TextCtrl: Diffusion-based Scene Text Editing with Prior Guidance Control (NeurIPS 2024 · similarity 0.88)
- Exploring the Deep Fusion of Large Language Models and Diffusion Transformers for Text-to-Image Synthesis (CVPR 2025 · similarity 0.88)
- Harmonizing Visual Text Comprehension and Generation (NeurIPS 2024 · similarity 0.87)
- SteerVTE: Seamless Video Text Editing with Style and Glyph Control (NeurIPS 2026 · similarity 0.87)
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.
