Integrity of Multimedia and Multimodal Data: From Capture to Use

With the development of Internet technologies, large volumes of multimedia and more generally, multimodal data can easily be shared among different users or entities for the purpose of diagnosis, analysis, and prediction in a variety of critical application contexts with no hard-copy backups being kept. This has serious implications for integrity, authentication, provenance, and utilization in a number of application domains. Additionally, various actions, including unlawful alteration, unauthorized distribution, copyright violation, and misuse of sensitive and private multimedia data, are growing each day producing negative impacts on the accuracy, efficiency, safety, and economics of multimodal data systems. Although there are many tools available to create, encrypt, embed, and extract data, there are relatively limited matured tools available and it remains a work-in-progress to establish integrity, authentication, and provenance. Therefore, addressing these challenges has been of interest for researchers in the field. Furthermore, various theories, techniques, and best practices developed by the community for handling multimedia data can provide useful inspirations or generalizable to multimodality data, including but not limited to multimedia. The advent of the Internet of Things (IoT), cyber-physical systems, robotics, as well as personal and wearable devices provides many opportunities for the multimedia community to develop synergies.