| --- |
| license: apache-2.0 |
| task_categories: |
| - image-classification |
| language: |
| - en |
| tags: |
| - AI-Generated Image Detection |
| - AI-Manipulated Images |
| size_categories: |
| - 100K<n<1M |
| --- |
| # DailyBench: A Unified Benchmark for AI-Generated and Manipulated Images from Modern Generative Models |
|
|
|
|
| <div align="center"> |
|
|
| [](https://dailybench.github.io/) |
| [](https://arxiv.org/abs/2607.24016) |
| [](https://huggingface.co/datasets/WhiteJiangzz/DailyBench) |
| [](https://modelscope.cn/datasets/WhiteJiang/DailyBench) |
|
|
| </div> |
|
|
| --- |
|
|
| ## Introduction |
|
|
| Recent advances in generative models have shifted AI-generated image detection from identifying easily distinguishable, fully synthetic images to identifying highly realistic content generated by both modern generation and manipulation pipelines. |
| However, existing detection benchmarks are often built with outdated generative models and primarily emphasize full-image synthesis, creating a growing mismatch between benchmark data and the images encountered in real-world generation and editing scenarios. |
| To bridge this gap, we introduce \textbf{DailyBench}, a high-quality unified benchmark for evaluating whether AI-generated image detectors can generalize across both modern full-image synthesis and object-level manipulation. |
| DailyBench contains two complementary subsets: \textbf{FakeBench}, which includes high-quality images synthesized by recent open-source and commercial generative models, and \textbf{ManipulationBench}, which introduces challenging object-level edits applied to real images using advanced image-conditional models. |
| This design makes DailyBench a realistic testbed for studying both generator-level generalization and manipulation-aware detection under subtle local edits. |
| In addition to the benchmark, we propose \textbf{FPD}~(\textbf{F}ake-\textbf{P}references \textbf{D}etector), a diagnostic-driven strong baseline that encourages diagnostic-sensitive representation learning. |
| FPD combines a dual-pathway architecture with a cascaded compression classifier to progressively extract and compact diagnostic-relevant cues, and further introduces a fake-preference sampler to reduce excessive bias toward the real class. |
| Experiments on DailyBench reveal substantial robustness gaps in current detectors: methods reporting 91-96\% balanced accuracy on GenImage drop to 60-79\% on FakeBench and 54-67\% on ManipulationBench. |
| These results show that existing detectors remain poorly generalized to realistic synthesis and manipulation, highlighting DailyBench as a rigorous testbed for developing robust and manipulation-aware AI-generated image detection methods. |
| The project is available at \url{https://dailybench.github.io/} |
|
|
|  |
|
|
| If this work is helpful for your research, please consider citing the following BibTeX entry. |
| ``` |
| @article{dailybench, |
| title={DailyBench: A Unified Benchmark for AI-Generated and Manipulated Images from Modern Generative Models}, |
| author={Xin Jiang, Hao Tang, Junyao Gao, Zijie Yang, Meiqi Cao, Fei Shen, Dongming Zhang, Zechao Li, Yongdong Zhang}, |
| journal={arXiv preprint arXiv:2607.24016}, |
| year={2026} |
| } |
