Table 2: Comparison with deep learning based methods for image forensics.

Method Main technique Application Hashing

Chen [15]Filter layer that output median filtering residualMedian Filtering detection No
Qian [16]Customized CNN modelSteganalysisNo
Bayar [17]New convolutional layer to learn manipulation detection featuresImage manipulation (i.e. median filtering, gaussian blurring, additive white Gaussian noise, eesampling) detection No

Bondi [18]Clustering of camera-based CNN featuresTampering Detection and Localization No
Yarlagadda [19]GAN and one-class classifierSatellite image forgery detection and LocalizationNo
D’Avino [20]Autoencoder with recurrent neural networksVideo forgery detection No

Tuama [21]A layer of preprocessing is added to the CNN modelCamera model identification No
Bondi [22]data-driven algorithm based on convolutional neural networksCamera model identification No

ProposedMultiview feature, perceptual saliency, semi-supervised hashingImage tamper detection Yes