Fully AI-Generated Image Detection: Definition, Recent Advances and Challenges

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Xu, Qijie, Wang, Can, Chen, Jiawei, Lyu, Siwei, Chen, Defang
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866910209997275136
author Xu, Qijie
Wang, Can
Chen, Jiawei
Lyu, Siwei
Chen, Defang
author_facet Xu, Qijie
Wang, Can
Chen, Jiawei
Lyu, Siwei
Chen, Defang
contents Recent advances in visual generative models have enabled the creation of highly realistic, fully AI-generated images without relying on real source content. While beneficial for many applications, these models also pose significant societal risks, as they can be easily exploited to produce convincing Deepfakes. Detecting them represents a foundational yet challenging problem in AI media forensics, requiring detectors to reliably extract the inherent artifacts imprinted by generative architectures. In this Review, we provide a systematic overview of fully AI-generated image detection. Following the standard detector design pipeline, we focus on two key components: dataset construction and artifact extraction. We analyze how dataset design influences the generalization and robustness of learned artifacts, and categorize existing artifact extraction methods based on the primary inductive priors leveraged to isolate artifacts. Within this framework, we systematically review existing works. Finally, we highlight open problems and envision several future directions for developing more robust and generalizable detectors. Reviewed works in this survey can be found at https://github.com/zju-pi/Awesome-Fully-AI-Generated-Image-Detection.
format Preprint
id arxiv_https___arxiv_org_abs_2502_19716
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Fully AI-Generated Image Detection: Definition, Recent Advances and Challenges
Xu, Qijie
Wang, Can
Chen, Jiawei
Lyu, Siwei
Chen, Defang
Computer Vision and Pattern Recognition
Machine Learning
Recent advances in visual generative models have enabled the creation of highly realistic, fully AI-generated images without relying on real source content. While beneficial for many applications, these models also pose significant societal risks, as they can be easily exploited to produce convincing Deepfakes. Detecting them represents a foundational yet challenging problem in AI media forensics, requiring detectors to reliably extract the inherent artifacts imprinted by generative architectures. In this Review, we provide a systematic overview of fully AI-generated image detection. Following the standard detector design pipeline, we focus on two key components: dataset construction and artifact extraction. We analyze how dataset design influences the generalization and robustness of learned artifacts, and categorize existing artifact extraction methods based on the primary inductive priors leveraged to isolate artifacts. Within this framework, we systematically review existing works. Finally, we highlight open problems and envision several future directions for developing more robust and generalizable detectors. Reviewed works in this survey can be found at https://github.com/zju-pi/Awesome-Fully-AI-Generated-Image-Detection.
title Fully AI-Generated Image Detection: Definition, Recent Advances and Challenges
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2502.19716