Skip to main content

Table 2 The performance of Linear Kernel-Net, Non-linear Kernel-Net, and Ye-Net on resampled images

From: Joint multi-domain feature learning for image steganalysis based on CNN

Algorithms

Linear Kernel-Net

Non-linear Kernel-Net

Ye-Net

WOW (0.2 bpp)

0.710

0.697

0.669

WOW (0.4 bpp)

0.819

0.788

0.768

S-UNIWARD (0.2 bpp)

0.661

0.609

0.600

S-UNIWARD (0.4 bpp)

0.766

0.734

0.688

  1. The involved networks are trained and tested on BOSSBase