Figure 7
From: FACSNet: Forensics aided content selection network for heterogeneous image steganalysis
Comparison of testing accuracy of FACSNet and the corresponding original SRNet. (a) The trained models are aiming at detecting HUGO spatial-domain stego images with 0.2/0.4 bpp payload. (b) The trained models are aiming at detecting J-UNIWARD stego images with 0.2/0.4bpnzAC payload and QF 95.
