![[2602.20951] See and Fix the Flaws: Enabling VLMs and Diffusion Models to Comprehend Visual Artifacts via Agentic Data Synthesis](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.20951] See and Fix the Flaws: Enabling VLMs and Diffusion Models to Comprehend Visual Artifacts via Agentic Data Synthesis
Summary
This paper presents ArtiAgent, a novel approach to automate the creation of artifact-annotated datasets for training visual language models and diffusion models, addressing the challenge of visual artifacts in AI-generated images.
Why It Matters
As AI-generated images become more prevalent, the presence of visual artifacts can undermine their realism. This research introduces an automated method to generate annotated datasets, which is crucial for improving the quality of AI outputs and advancing the field of computer vision.
Key Takeaways
- ArtiAgent automates the generation of artifact-annotated datasets.
- The approach includes three agents: perception, synthesis, and curation.
- The method synthesizes 100K images with rich artifact annotations.
- It addresses the limitations of human-labeled datasets in scalability.
- The research demonstrates efficacy across diverse applications.
Computer Science > Computer Vision and Pattern Recognition arXiv:2602.20951 (cs) [Submitted on 24 Feb 2026] Title:See and Fix the Flaws: Enabling VLMs and Diffusion Models to Comprehend Visual Artifacts via Agentic Data Synthesis Authors:Jaehyun Park, Minyoung Ahn, Minkyu Kim, Jonghyun Lee, Jae-Gil Lee, Dongmin Park View a PDF of the paper titled See and Fix the Flaws: Enabling VLMs and Diffusion Models to Comprehend Visual Artifacts via Agentic Data Synthesis, by Jaehyun Park and 5 other authors View PDF Abstract:Despite recent advances in diffusion models, AI generated images still often contain visual artifacts that compromise realism. Although more thorough pre-training and bigger models might reduce artifacts, there is no assurance that they can be completely eliminated, which makes artifact mitigation a highly crucial area of study. Previous artifact-aware methodologies depend on human-labeled artifact datasets, which are costly and difficult to scale, underscoring the need for an automated approach to reliably acquire artifact-annotated datasets. In this paper, we propose ArtiAgent, which efficiently creates pairs of real and artifact-injected images. It comprises three agents: a perception agent that recognizes and grounds entities and subentities from real images, a synthesis agent that introduces artifacts via artifact injection tools through novel patch-wise embedding manipulation within a diffusion transformer, and a curation agent that filters the synthesized ...
