AI Paper Atlas

Research map / Image generation / Text-to-image diffusion

TextAtlas5M: A Large-Scale Dataset for Long Text Image Generation

Dongxing Mao, Alex Jinpeng Wang, Weiming Han, Jiawei Zhang, Zhuobai Dong, Linjie Li, Yiqi Lin, Zhengyuan Yang, Libo Qin, Fuwei Zhang, Lijuan Wang, Min Li

ICML 2026 · Image generation › Text-to-image diffusion · approach: Personalized and region-controlled generation

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 changedWhereMeasuredEffect
fine-tune Lumina-mGPT 7B (1024) on TextAtlas5MLumina-mGPT 7B · TextAtlasEval CleanTextSynthOCR accuracy: 0.32 -> 48.55Improved
fine-tune Lumina-mGPT 7B (1024) on TextAtlas5MLumina-mGPT 7B · TextAtlasEval TextVisionBlendOCR accuracy: 0.03 -> 65.34Improved
fine-tune PixArt-alpha 0.6B (512) on TextAtlas5MPixArt-alpha 0.6B · TextAtlasEval CleanTextSynthOCR accuracy: 0.23 -> 39.14Improved
fine-tune Janus-Pro 1B on TextAtlas5MJanus-Pro 1B · TextAtlasEval CleanTextSynthOCR accuracy: 0.27 -> 5.23Improved
fine-tune on TextAtlas5M, hard real-image subsetLumina-mGPT 7B (1024) · TextAtlasEval TextScenesHQOCR accuracy: 0.22 -> 6.44Improved
train on TextAtlas5M instead of Marion10M (ablation)PixArt-alpha 0.6B (512) · LAIONEval4000 (short text)OCR accuracy: 52.37 -> 67.53Improved
add synthetic subsets to real-only training data (ablation)Lumina-mGPT 7B · TextScenesHQ EvalOCR accuracy: 5.32 -> 6.44Improved
raise Lumina-mGPT resolution from 512 to 1024 after fine-tuning (ablation)Lumina-mGPT 7B fine-tuned · TextAtlasEval CleanTextSynthOCR accuracy: 32.90 -> 48.55Improved
fine-tune Lumina-mGPT 7B (512) on TextAtlas5MLumina-mGPT 7B (512) · TextAtlasEval TextVisionBlend CLIP scoreCLIP score: 0.19 -> 0.16Hurt
fine-tune on TextAtlas5M, StyledTextSynth subsetLumina-mGPT 7B (512) · TextAtlasEval StyledTextSynthOCR accuracy: 0.09 -> 5.43Improved
use lower-resolution discrete VQ tokenizer for AR text rendering (ablation)Janus-Pro VQ-VAE tokenizer · image reconstructiontext fidelityHurt
increase input resolution of VQ-VAE tokenizer (256 to 1024) (ablation)Janus-Pro VQ-VAE tokenizer · image reconstructiontext clarityImproved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy