Guess the unified model: Domain and Linguistic Effects in Generated Images

Jasin Cekinmez, Ryo Mitsuhashi, Yida Yin

Keywords: Data Attribution, Synthetic Image Provenance, Unified Models, Linguistic Separability, Computer Vision

TL;DR: Images generated by different unified models are surprisingly easy to attribute to their source model even under corruptions and across domains, while prompt language leaves little to no reliable visual signature.

Abstract: With unified model-generated images now widespread online, attributing their model of origin offers a path toward transparency and deeper insight into the characteristic behaviors of individual models. Prior work has explored provenance in LLM generated text, diffusion model images, and datasets, but the separability of unified model-generated images remains an underexplored area. We address this gap by examining separability across corruption, domains, and prompt languages using images generated by seven unified models. We show that model attribution is highly feasible as our model achieves near-perfect accuracy with around 20K images per model. Corruptions and structural perturbations have only a modest effect on attribution performance, and cross-domain generalization reveals that semantic content contributes to separability but is not the dominant signal. Finally, we observe that for most models, prompt language attribution is around chance levels, suggesting minimal language specific visual signatures. These findings highlight consistent model-specific visual characteristics in unified models outputs and open new directions for tracing, auditing, and securing generative image pipelines.

Email Sharing: We authorize the sharing of all author emails with Program Chairs.

Data Release: We authorize the release of our submission and author names to the public in the event of acceptance.

Submission Number: 29