TIQA: Human-Aligned Perceptual Text Quality Assessment in Generated Images
Not on the map. Why?
Abstract (arXiv)
Recent text-to-image models have improved global realism, but text rendering remains a persistent failure mode: images may look convincing overall, yet local typography often contains malformed glyphs, broken strokes, irregular spacing, and other artifacts that humans heavily penalize. We formulate Text-in-Image Quality Assessment (TIQA), a no-reference task that estimates a human-aligned perceptual quality score for detected text regions while disentangling visual text quality from semantic correctness. To support this setting, we introduce two datasets. TIQA-Crops contains 120k text crops from 36k AI-generated images produced by 12 generators, with 10k mean-opinion-score (MOS) labels and 110k proxy labels for pretraining. TIQA-Images contains 1,500 text-heavy images from 10 recent generators, including proprietary systems, with paired overall-quality and text-quality subjective scores. We also propose ANTIQA, a lightweight predictor with text-specific inductive biases. Across crop-level and image-level evaluations, ANTIQA achieves the best alignment with human judgments, reaching PLCC/SROCC of 0.942/0.935 on TIQA-Crops and 0.842/0.837 for text-quality MOS on unseen generators in TIQA-Images. In best-of-5 AI-generated image ranking, ANTIQA improves the text quality of the selected image by 0.36 MOS (14%), demonstrating utility for benchmarking, filtering, and generation-time selection. Together, these findings establish perceptual text quality as a distinct evaluation target for modern text-to-image generation.
Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Builds on 22 of its 57 references that are in the map. First posted on arXiv 7 Mar 2026.
Builds on
22 of 57 references.
References · 22
- 2015Deep Residual Learning for Image Recognitioncited by 52,096
- 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,186
- 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,177
- 2017Squeeze-and-Excitation Networkscited by 4,789
- 2022LAION-5B: An open large-scale dataset for training next generation image-text modelscited by 4,152
- 2016Improved Techniques for Training GANscited by 2,970
- 2024A Survey on LLM-as-a-Judgecited by 1,423
- 2023Pick-a-Pic: An Open Dataset of User Preferences for Text-to-Image Generationcited by 977
- 2025GLM-4.5V and GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learningcited by 466
- 2017NIMA: Neural Image Assessmentcited by 389
- 2023TextDiffuser: Diffusion Models as Text Painterscited by 229
- 2023TOPIQ: A Top-down Approach from Semantics to Distortions for Image Quality Assessmentcited by 226
10 more
- 2023AnyText: Multilingual Visual Text Generation And Editingcited by 170
- 2025Teaching Large Language Models to Regress Accurate Image Quality Scores using Score Distributioncited by 141
- 2018Show, Attend and Read: A Simple and Strong Baseline for Irregular Text Recognitioncited by 130
- 2023Depicting Beyond Scores: Advancing Image Quality Assessment through Multi-modal Language Modelscited by 105
- 2024Adaptive Image Quality Assessment via Teaching Large Multimodal Model to Comparecited by 66
- 2025Q-Eval-100K: Evaluating Visual Quality and Alignment Level for Text-to-Vision Contentcited by 32
- 2024Type-R: Automatically Retouching Typos for Text-to-Image Generationcited by 7
- 2025STRICT: Stress Test of Rendering Images Containing Textcited by 7
- 2025TextInVision: Text and Prompt Complexity Driven Visual Text Generation Benchmarkcited by 5
- 2024TypeScore: A Text Fidelity Metric for Text-to-Image Generative Modelscited by 4
Built on by
Citing papers per year: 2026: 1.
1 paper.