IMAGE AND VIDEO RETRIEVAL AND AUTHENTICATION USING AI-DRIVEN TECHNIQUES FOR SECURE MEDIA MANAGEMENT

The proliferation of digital media has led to an increased need for secure and efficient systems for image and video retrieval and authentication. Traditional approaches often struggle with scalability and vulnerability to tampering, compromising the integrity of media management systems. The rise of artificial intelligence and deep neural networks (DNNs) offers transformative potential to address these challenges. By leveraging DNNs, this study proposes an advanced framework for secure media management, integrating robust retrieval and authentication mechanisms. The method employs a convolutional neural network (CNN)-based encoder-decoder architecture to extract and match high-dimensional features for image and video retrieval. For authentication, a blockchain-backed hash validation ensures the originality and integrity of media assets. The system is trained and evaluated on benchmark datasets, such as MS-COCO and UCF101, with augmentation techniques enhancing its adaptability across diverse media formats and resolutions. Key performance metrics include retrieval accuracy, processing time, and authentication robustness. Experimental results show a retrieval accuracy of 96.8%, with a mean processing time of 0.85 seconds per query. Authentication robustness achieves a 99.2% success rate in detecting altered media, significantly outperforming existing systems. The proposed framework ensures both scalability and security, offering an innovative solution for media management in domains such as journalism, legal evidence management, and social media platforms.