See and Fix the Flaws: Enabling VLMs and Diffusion Models to Comprehend Visual Artifacts via Agentic Data Synthesis

TL;DR

This paper introduces ArtiAgent, an automated system that synthesizes and annotates images with visual artifacts to improve the training and evaluation of vision-language models and diffusion models.

Contribution

ArtiAgent automates the creation of large-scale artifact-annotated datasets, reducing reliance on costly human labeling and enabling scalable artifact mitigation in image generation models.

Findings

Synthesized 100K images with artifact annotations

Demonstrated effectiveness across diverse applications

Enabled improved artifact detection and mitigation

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 artifacts and generates both local and global explanations for each instance. Using ArtiAgent, we synthesize 100K images with rich artifact annotations and demonstrate both efficacy and versatility across diverse applications. Code is available at link.