Image Steganography Using Auto Encoder-Decoder Based Deep Learning Method

Image steganography is the process of communicating hidden secret image embedded in a cover image in plain sight without arousing any suspicions. In the recent times, deep learning methods have gained popularity and is widely used in the field of steganography. In this paper, an auto encoder-decoder based deep convolutional neural network is proposed to embed the secret image inside the cover image and to extract the secret image from the generated stego image. Training and testing is done on a subset of the ImageNet dataset. To evaluate the proposed method, Peak Signal-Noise Ratio (PSNR) and Mean Squared Error (MSE) metrics are used. The proposed method has proved to achieve higher invisibility, security and robustness. The capacity of the method is higher when compared to traditional Least Significant Bit substitution methods.