Fake Generated Painting Detection via Frequency Analysis
This work addresses the detection of fake paintings generated by style transfer algorithms, which is an incremental improvement for digital art authentication.
The paper tackles the problem of detecting digitally generated fake paintings by analyzing statistical differences in the Fourier frequency domain, proposing a method that extracts three types of frequency features and demonstrates excellence in various testing conditions.
With the development of deep neural networks, digital fake paintings can be generated by various style transfer algorithms.To detect the fake generated paintings, we analyze the fake generated and real paintings in Fourier frequency domain and observe statistical differences and artifacts. Based on our observations, we propose Fake Generated Painting Detection via Frequency Analysis (FGPD-FA) by extracting three types of features in frequency domain. Besides, we also propose a digital fake painting detection database for assessing the proposed method. Experimental results demonstrate the excellence of the proposed method in different testing conditions.