Enhanced CNN Approaches for Multi-Image Embedding in Image Steganography

This paper introduces advanced convolutional neural network (CNN) techniques designed to improve multi-image embedding in image steganography. By leveraging deep learning, our approach achieves higher data hiding capacity while maintaining imperceptibility and robustness against detection. We propose a novel architecture that optimizes the embedding process, enabling the concealment of multiple images within a single cover image. Experimental results demonstrate superior performance in terms of payload capacity, visual quality, and security compared to existing steganography methods.

Recommended citation: Md. Irtiza Hossain, Samiul Kadir, Farhan Ishraq Fagun, Ishtiaq Samiul, Rafi Zaman Saukhin. (2024). "Enhanced CNN Approaches for Multi-Image Embedding in Image Steganography." 2024 International Conference on Information and Communication Technologies (ICICT).
Download Paper