Image Steganography using Encoder - Decoder Architectures

Network security and data protection lie at the very core of everything technology - from messaging apps to IOT devices. It is fundamental that all users send out and receive information intended only for them with cybersecurity best practices guarding their information at all times. A substantial aspect of cybersecurity is data protection which is brought to life via encryption. Encryption protects private information and sensitive data and is essential to everyday communication on the Internet. Steganography and Cryptography are important methods that are used to enhance the process of encryption. While effective, encryption is dependent on keys which if compromised can lead to data getting in unauthorised hands and potentially getting misused. Additionally, with the advent of quantum technologies there is a conversation around the possibility of quantum algorithms being able to break through the highest forms of encryption. This has inspired the research on using neural networks to implement data protection which can effectively protect data without a dependency on keys. This paper aims to research and compare various techniques of performing data protection using image steganography through encoder-decoder architectures which are a special type of neural networks and to identify aspects from reviewed papers that will form basis for implementation of the same.