T2I Generation Evaluation

T2I: Text-to-Image

Follow this topic Ask about this topic

Momentum

9 papers in the last four weeks, up 50% on the four weeks before. 0.1% of all new papers.

Jul 13Week of Sep 28

Latest papers 98

All topics

It's Always 10:10: Reference Images Break a Bias That Prompts Only Dent
Oct 8, 2026Luca CazzanigaT2I Generation T2I Generation Evaluation
Independent Researcher, AI-Assisted Visual Production
When Scene Text Hijacks the Scene: Uncovering, Exploiting, and Mitigating Rendered-Text Semantic Leakage in Image Generation Models
Oct 8, 2026Feifei Li, Runjie Wang, Xiaohan Zhang +3Safe T2I Generation Prompt Injection Attacks
Fudan University · Shanghai University of Electric Power
False Claims, Credible Images: A Red-Teaming Benchmark for Commercial Image Generators
Oct 8, 2026Zeyu Ye, Yanchun Li, Sibei He +8Safe T2I Generation T2I Generation Evaluation
Xiangtan University · Jinan University · Huazhong University of Science and Technology +4
UltraText Bench: A Comprehensive Bilingual Benchmark for Evaluating Visual Text Rendering in Image Generation
Oct 7, 2026Deyuan Liu, Yihao Hu, Jingxuan Zhang +13T2I Generation Evaluation T2I Generation
Westlake University · Ant Group · Zhejiang University +7
Harmful Content Generation in Text-to-Image Models: Capabilities and Moderation Limitations
Oct 5, 2026Paschalis Giakoumoglou, Manos Schinas, Symeon PapadopoulosT2I Generation Content Moderation
Centre for Research and Technology Hellas (CERTH), Thessaloniki, Greece
Evaluating the Evaluators: Diagnosing Large Multimodal Models for AI-Generated Image Assessment
Sep 29, 2026Yu Zhao, Jiarui Wang, Huiyu Duan +5LLM-as-a-Judge Image Generation Evaluation
Shanghai Jiao Tong University · Dalian University of Technology
Verifiable Visual Rewards Transfer from Synthetic Scenes to Natural Prompts
Sep 28, 2026Shuyue Stella Li, Xiaochuang Han, Yulia Tsvetkov +1Compositional T2I Generation RL for Image Generation
University of Washington
eval-unlearn: Benchmarking unlearning in Text-to-Image Diffusion Models
Sep 28, 2026Mansi, Nikhil Raghavan, Zixia Huang +4Diffusion Model Unlearning T2I Generation Evaluation
Imperial College London
RISE: Red-teaming via Iterative Strategy Evolution for Modern Text-to-Image Models
Sep 28, 2026Dmitrii Kharlapenko, Sergei Bratchikov, Konstantin Korolev +1Adversarial Prompt Generation T2I Generation Evaluation
White Circle
InGuard: Towards Generalized Inner Guardrail for Safe Text-to-Image Generation
Sep 23, 2026Zeyu Wang, Xiaodan Li, Zhiwen Li +2T2I Generation Safe T2I Generation
Alibaba AAIG
IMPLICIT-Bench: Measuring Implicit Bias in Text-to-Image Models under Neutral Prompts
Sep 21, 2026Yue Dai, Ziyang Liu, Marc Cheong +1T2I Generation Demographic Bias in T2I Models
The University of Melbourne
Newer Is Not Fairer: Gender Stereotyping in Text-to-Image AI Across Model Generations
Sep 16, 2026Shesh Narayan Gupta, Nik Bear BrownT2I Generation Demographic Bias in T2I Models
College of Engineering, Northeastern University, Boston, MA 02115, USA
Certifying Concept Unlearning in Text-to-Image Diffusion Models
Sep 14, 2026Mansi, Luca Marzari, Francesco LeofanteDiffusion Model Unlearning Concept Erasure
Imperial College London · TU Wien
Prompt Revision as a Source of Cultural Bias in Text-to-Image Systems
Sep 12, 2026Aleksandra Urman, Elsa Lichtenegger, Salima Jaoua +7T2I Generation Cultural Bias in Language Models
University of Zurich · University of Helsinki · ETH Zurich +1
Blending Concepts: Benchmarking Visual Metaphor Generation in Text-to-Image Models
Sep 2, 2026Chuer Chen, Zichen Wang, Yi He +2T2I Generation T2I Generation Evaluation
1Tongji University · 2Shanghai Innovation Institute
ImageEval 2026: Culturally Grounded Arabic Multimodal Evaluation
Aug 31, 2026Samir Abdaljalil, Hunzalah Hassan Bhatti, Ahlam Bashiti +11Hallucination Detection T2I Generation
Texas A&M University · Qatar Computing Research Institute, Qatar · Birzeit University, Palestine +2
ContextBias: Controlled Evaluation of Bias Persistence Under Context Shift in Text-to-Image Models
Aug 30, 2026Shaghayegh Kolli, Sina Emami, Moreno D'Incà +4Demographic Bias in T2I Models T2I Generation Evaluation
1Technical University of Munich · 4Munich Center for Machine Learning (MCML) · 5Munich Data Science Institute (MDSI) +2
TangPoetryBench: A Multi-Dimensional Benchmark and Rubric-Conditioned Evaluator for Poetry-to-Image Generation
Aug 11, 2026Haoqi Hu, Tongji Luo, Li Zhang +1Image Generation Automated Evaluation
Independent Researcher
On the Limitations of Cross-Lingual Consistency in Multilingual Text-to-image Generation
Aug 11, 2026Sicheng Zhang, Zhonghao Yan, Binzhu Xie +4T2I Generation T2I Generation Evaluation
Khalifa University · Queen Mary University of London · The Chinese University of Hong Kong +3
Open Evaluation Agent: Efficient and Promptable Evaluation of Visual Generative Models
Aug 10, 2026Shulin Tian, Ziqi Huang, Fan Zhang +3Tool-Augmented Language Model Agents AI Agent Evaluation
Shanghai Artificial Intelligence Laboratory
Math-Vision Diagrams: A Comprehensive Benchmark for Evaluating LLM Mathematical Diagram Generation Capabilities
Aug 9, 2026Harish Kashyap, Kiran Byadarhaly, Sriram Chakaravarthy +2LLM Evaluation T2I Generation Evaluation
Pandita AI Inc., 2540 N First St, San Jose CA 95131, USA
Simile Understanding in Text-to-Image Models: An Evaluation Framework
Aug 5, 2026Luecheng Wang, Shintaro Ozaki, Hidetaka Kamigaito +4Figurative Language Understanding Vision-Language Grounding
The University of Tokyo · Nara Institute of Science and Technology · Chungnam National University +1
MultiCompose: Multi-Concept Personalized Composition with Per-Subject Attribute Binding
Aug 4, 2026Ruirui Zhang, Zhengkai Zhao, Pan GaoImage Generation Compositional T2I Generation
Nanjing University of Aeronautics and Astronautics College of Computer Science and Technology Nanjing, Jiangsu, China · Nanjing University of Aeronautics and Astronautics College of Artificial Intelligence Nanjing, Jiangsu, China
Can Text-to-Image Models Draw from the Right Frame of Reference?
Aug 4, 2026Zheyuan Gu, Ruihang Li, Yong Huang +5Vision-Language Models T2I Generation
1ERNIE Team, Baidu Inc. · 2Peking University
Investigating Social Bias in Narrative Image Generation
Aug 3, 2026Junyeong Park, Sowon Min, Euna Jang +5T2I Generation Demographic Bias in T2I Models
KAIST, South Korea
NuclearDiffusion: Text-to-Image Foundation Models for Learning Nuclear Energy Concepts
Aug 1, 2026Mohammed I. Radaideh, Jeremy Moon, Andre Gala-Garza +3Domain Adaptation Diffusion Model Fine-Tuning
Department of Mechanical Engineering, University of Michigan, Ann Arbor, Michigan 48109, USA · Michigan Institute for Computational Discovery and Engineering (MICDE), University of Michigan, Ann Arbor, Michigan 48109, USA · Department of Electrical Engineering and Computer Science, University of Michigan, Ann Arbor, Michigan 48109, USA +1
Scaling Properties of Text Conditioning in Visual Generation
Jul 31, 2026Zilong Chen, Chaorui Deng, Kunchang Li +2T2I Generation Compositional T2I Generation
ByteDance Seed
OmniPhys: Knowledge-Graph-Driven Benchmarking and Collective Optimization for Physical Commonsense in Text-to-Image Generation
Jul 28, 2026Yajing Xu, Yarong Lan, Jiaoyan Chen +6T2I Generation Prompt Optimization for T2I Generation
Zhejiang University · The University of Manchester · The University of Edinburgh +1
LU-500: A Logo Benchmark for Concept Unlearning
Jul 27, 2026Keyu Li, Jin Gao, Jialing Zhang +1Diffusion Model Unlearning Concept Erasure
Shanghai Jiao Tong University Shanghai, China
ExpertVerse: A General-Purpose Benchmark for Expert-Level Reasoning in Knowledge-Intensive Visual Synthesis
Jul 21, 2026Yuan Wang, Yongchao Du, Mengting Chen +4Image Generation RL for Image Generation
1Alibaba Group · 2Tsinghua University · 3Johns Hopkins University‌

Oct 8, 2026, cs.CVJ/K move · Enter open · S save

It's Always 10:10: Reference Images Break a Bias That Prompts Only Dent

Luca Cazzaniga

Independent Researcher, AI-Assisted Visual Production

T2I Generation T2I Generation Evaluation

Open paper AskPDF arXiv

Text-to-image models appear to reproduce the habits of the photographs they learned from. Analog clocks are an extreme case: in advertising, watches almost always show 10:10, and generated clocks return to 10:10 even when another time is requested. We measure this bias and test three ways of overcoming it on 52 models available on the Magnific platform, with a replication on Higgsfield. Every image shows three identical clocks that must show 2:35, 6:50 and 11:20. The description of the object is fixed and only the request about the time changes: no time (A), the time in digits (B), the hand positions described by construction relative to the dial numerals (C), or the same description plus a drawn reference dial (D). Two AI readers read all 1,799 images blind from coded copies, with a third reader and the author settling disagreements (dial-level agreement 96.0% and 97.3%). With no time requested, 67% of the images have all three clocks at 10:10. On the 20 current models, all three clocks are correct in 34% of the images with digits, 30% with the hands described in words and 75% with the reference dial (D-B: +37 points, 95% CI +28 to +45); we found no evidence that describing the hands in words beats the digits (C-B: -4 points, CI -10 to +1). The replication on the 12 models shared by both platforms gives the same picture (B 54%, C 50%, D 81%). Writing the time reduces the bias but leaves two thirds of the images of the 20 current models with at least one wrong clock; adding a drawn reference raises full accuracy to three quarters and almost eliminates images entirely at 10:10. We release all images, prompts, raw readings and a script that recomputes every result.

Momentum

Related topics

Latest papers 98