IntrAA-SD：面向 Stable Diffusion 的同架构微调溯源新任务与基准数据集
Intra-Architecture Source Attribution for Stable Diffusion: A New Task with Benchmark Dataset and Baseline
AI 导读
研究者提出新任务 IntrAA-SD，针对 Stable Diffusion 同架构、仅因微调而不同的生成图溯源，并发布首个包含 140K 张图像的基准数据集，涵盖 SDXL 各微调版本及域内和分布外提示词。基线方法采用冻结 CLIP 主干捕获风格并门控 MoE，结合高分辨率 LoRA 适配的 DINOv2 提取微调痕迹特征。该工作已被 WACV 2027 接收，代码与数据集已开源。
正文 · 原文
Abstract:Designing systems that attribute a generated image to its source model has lately become a major area of AI research. Most existing datasets and systems related to image source attribution are targeted towards the detection of architectural signatures of the source models. Distinct from this, focusing on stable diffusion, we introduce a novel task, IntrAA-SD, Intra-Architecture source Attribution for Stable Diffusion, where the source models to be differentiated share the same architectural backbone and differ only due to fine-tuning. The proposed task aligns with the emergence of fine-tuned generators having the same architecture, like Juggernaut XL and RealVis XL, which are fine-tuned versions of Stable Diffusion XL (SDXL). We provide a first-of-its-kind $140$K-image benchmark dataset containing images generated from different fine-tunes of SDXL, along with in-domain and out-of-distribution prompts. We also provide a baseline framework for IntrAA-SD, which leverages a style-conditioned Mixture-of-Experts network. It uses a frozen CLIP backbone to capture the image style and gate the MoE, which processes fine-tuned trace features from a high-resolution LoRA-adapted DINOv2 network. The fine-tuned models used, the prompting strategies considered, and the formation details employed in the novel dataset are elaborated. Our baseline intra-architecture source attribution method is evaluated in terms of attribution accuracy on our dataset, and ablation studies are also performed. Code and dataset is available at this https URL.
| Comments: | Accepted at IEEE/CVF Winter Conference on Applications of Computer Vision(WACV 2027) |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.06000 [cs.CV] |
| (or arXiv:2611.06000v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2611.06000 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Mallika Garg [view email]
[v1]
Wed, 4 Nov 2026 05:47:50 UTC (27,826 KB)
来源：arXiv cs.AI · arxiv.org