Search for a command to run...
In an era where digital security is paramount, ensuring the confidentiality of sensitive images across diverse applications—ranging from personal data protection to medical imaging and secure communications—is crucial. Traditional encryption methods face challenges in balancing security, efficiency, and robustness against adversarial attacks. This paper explores a novel approach to image encryption and decryption using Graph Neural Networks (GNNs). GNNs leverage the inherent spatial and structural dependencies of image data, enabling adaptive encryption schemes that enhance security while maintaining computational efficiency. Our proposed framework transforms image pixels into graph representations, applies complex transformations through graph convolutional layers, and reconstructs the original image upon decryption with minimal loss. The images recovered through the proposed decryption technique are of high quality, as justified by PSNR and SSIM values. To validate the proposed framework, we demonstrate its robustness against a representative cryptographic attack, highlighting its resistance to unauthorized access and its computational feasibility for practical deployment in secure communication systems. To the best of our knowledge, this is the first work that integrates GNN techniques for both image encryption and decryption, offering a more secure and resilient approach compared to state-of-the-art methods. This research opens new avenues for secure image transmission in a wide range of sensitive domains, ensuring privacy and confidentiality through advanced machine learning techniques.