Sign inGet started
Dataset · Hugging Face

yuzj1997yzj/WildAIGC-Bench

WildAIGC-Bench AI images behind real news events and online claims. A hard test for today’s detectors. Paper on arXiv · GitHub repository · Project page · Evaluation results Broad Coverage · Timely Updates · Human-Verified Labels WildAIGC-Bench evaluates detection of photographic-looking AI-generated and edited images encountered in real online circulation. The current v1 release contains 407 images spanning diverse events and forms of generation and editing, with acquired and… See the full description on the dataset page: https://huggingface.co/datasets/yuzj1997yzj/WildAIGC-Bench.

Author / organizationyuzj1997yzj
License (as declared)other
Downloads157 (as reported by Hugging Face)
Likes0
Created10 October 2026
Last updated at source21 October 2026
imageai-generated-image-detectionin-the-wildbenchmark
Embed this entry
<iframe src="https://silytics.si/embed/item/49890c25-97ac-45d0-8bb7-26d4c32512e6" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Connected in the knowledge graph

Explore this entry in the graph

Related papers

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 21 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.