ai-generated-image-detection
AI-generated-Image-DetectionA standardized image dataset (150 images total) was built under the principle of same semantics with different sources, controlling semantic and visual consistency to avoid irrelevant variable interference. It included 100 AI-generated images (Doubao, ByteDance Inc., China; Kimi, Moonshot AI Inc., China) and 50 real photography images sourced from Pexels (https://www.pexels.com/zh-cn/), covering five categories: natural scenery(zrfg), portrait(rwxx), animals and plants(dzw), food still… See the full description on the dataset page: https://huggingface.co/datasets/JingmingWang/AI-generated-Image-Detection.AI-generated-Image-DetectionA standardized image dataset (150 images total) was built under the principle of same semantics with different sources, controlling semantic and visual consistency to avoid irrelevant variable interference. It included 100 AI-generated images (Doubao, ByteDance Inc., China; Kimi, Moonshot AI Inc., China) and 50 real photography images sourced from Pexels (https://www.pexels.com/zh-cn/), covering five categories: natural scenery(zrfg), portrait(rwxx), animals and plants(dzw), food still… See the full description on the dataset page: https://huggingface.co/datasets/whikt/AI-generated-Image-Detection.CAS3120-2026-Project1_AI-Generated-Image-Detection
Project 1 — AI-Generated Image Detection
Course materials for CAS3120 · Introduction to Machine Learning · Spring 2026, Department of AI, Yonsei University.
Task
Binary image classification: distinguish real images from AI-generated images.
Dataset Summary
Image size: 128 × 128 RGB PNG
Splits:
train: 2,000 images (labeled)
val: 1,000 images (labeled)
test: 2,000 images (labels withheld)
Class balance: 50/50 in each labeled split
Test labels are withheld.… See the full description on the dataset page: https://huggingface.co/datasets/yonsei-dli/CAS3120-2026-Project1_AI-Generated-Image-Detection.Robustness-of-AI-Generated-Image-Detection-Against-Localized-Inpainting-Attacks
Robustness of AI-Generated Image Detection Against Localized Inpainting Attacks
This repository hosts the detector-ready datasets and mask packs used in the thesis:
Robustness of AI-Generated Image Detection Against Localized Inpainting AttacksOguz Akin, Saarland University, CISPA Helmholtz Center for Information Security (2025)
It provides standardized evaluation splits for six state-of-the-art AI-generated image (AIGI) detectors across watermarking, passive, and training-free… See the full description on the dataset page: https://huggingface.co/datasets/eoguzakin/Robustness-of-AI-Generated-Image-Detection-Against-Localized-Inpainting-Attacks.Project1-AI-Generated-Image-Detection-2026
Project 1 — AI-Generated Image Detection
Course materials for CAS3120 · Introduction to Machine Learning · Spring 2026, Department of AI, Yonsei University.
Task
Binary image classification: distinguish real images from AI-generated images.
Dataset Summary
Image size: 128 × 128 RGB PNG
Splits:
train: 2,000 images (labeled)
val: 1,000 images (labeled)
test: 2,000 images (labels withheld)
Class balance: 50/50 in each labeled split
Test labels are withheld.… See the full description on the dataset page: https://huggingface.co/datasets/legenduck/Project1-AI-Generated-Image-Detection-2026.AI-image-detection-AI-generated