Using Neural Networks for Fake Colorized Image Detection - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2019

Using Neural Networks for Fake Colorized Image Detection

Yuze Li
  • Function : Author
Yaping Zhang
  • Function : Author
  • PersonId : 1067594
Liangfu Lu
  • Function : Author
Yongheng Jia
  • Function : Author
Jingcheng Liu
  • Function : Author


Modern colorization techniques can create artificially-colorized images that are indistinguishable from natural color images. As a result, the detection of fake colorized images is attracting the interest of the digital forensics research community. This chapter tackles the challenge by introducing a detection approach that leverages neural networks. It analyzes the statistical differences between fake colorized images and their corresponding natural images, and shows that significant differences exist. A simple, but effective, feature extraction technique is proposed that utilizes cosine similarity to measure the overall similarity of normalized histogram distributions of various channels for natural and fake images. A special neural network with a simple structure but good performance is trained to detect fake colorized images. Experiments with datasets containing fake colorized images generated by three state-of-the-art colorization techniques demonstrate the performance and robustness of the proposed approach.
Fichier principal
Vignette du fichier
488399_1_En_11_Chapter.pdf (161.86 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-02534604 , version 1 (07-04-2020)





Yuze Li, Yaping Zhang, Liangfu Lu, Yongheng Jia, Jingcheng Liu. Using Neural Networks for Fake Colorized Image Detection. 15th IFIP International Conference on Digital Forensics (DigitalForensics), Jan 2019, Orlando, FL, United States. pp.201-215, ⟨10.1007/978-3-030-28752-8_11⟩. ⟨hal-02534604⟩
53 View
9 Download



Gmail Facebook Twitter LinkedIn More