Toward High Capacity and Robust JPEG Steganography Based on Adversarial Training
Algorithm 1
Training steps of HRJS.
Input: Secret message;
DCT coefficients of the cover image.
Step 1: Connect the DCT coefficients of the cover image with a binary secret message and input them into the encoder to obtain the DCT coefficients of the stego image;
Step 2: Convert the DCT coefficients of cover and stego images into corresponding spatial images by utilizing the IDCT module;
Step 3: Input the spatial stego image into the attack module to obtain the attacked stego image;
Step 4: Convert the attacked stego image into DCT coefficients;
Step 5: Input the DCT coefficients of attacked stego image into the decoder to obtain the decoded secret message;
Step 6: Feed spatial cover and stego image into the discriminator;
