Steganographic Network
Steganographic networks leverage deep learning to hide secret data within seemingly innocuous cover media, such as images, aiming for undetectable covert communication. Current research focuses on developing efficient and robust network architectures, including those that disguise themselves as ordinary machine learning models or leverage techniques like style transfer to mask the embedded information, and on optimizing these networks for size and computational efficiency. This field is significant for its potential applications in secure communication and data hiding, while also driving advancements in model compression and deep learning architecture search within the broader context of steganography and steganalysis.
Papers
February 27, 2024
February 28, 2023
June 12, 2022