Modeling and Optimization of Signals Using Machine Learning Techniques

17Current Advancements in Steganography: A Review

Mallika Garg^1*, Jagpal Singh Ubhi² and Ashwani Kumar Aggarwal³

¹Department of Electronics and Communication Engineering, Indian Institute of Technology, Roorkee, Uttarakhand, India

²Electronics and Communication Engineering Department, Sant Longowal Institute of Engineering and Technology, Longowal, Punjab, India

³Electrical and Instrumentation Engineering Department, Sant Longowal Institute of Engineering and Technology, Longowal, Punjab, India

Abstract

In today’s world, communication over the Internet has increased to a great extent. With this increased communication, enormous data are to be kept and communicated secretly and safely. Many techniques have been developed to ensure the secret transfer of data over the internet. Steganography is used to protect data from unauthorized persons. This chapter provides a review of various traditional, convolutional neural network (CNN)-, and generative adversarial network (GAN)- based steganographic techniques. Details of the different types of steganography based on the host and domain, its applications, its disadvantages, and various evaluation parameters are also illustrated. This paper also presents the details of the datasets used for steganography and contains some basic details of the counterpart of steganography called steganalysis.

Keywords: Image security, convolutional neural network, information hiding, StegoExpose, steganography

17.1 Introduction

Steganography’s primary ...