The proliferation of imaging and storage devices and the ubiquitous presence of computer networks make sharing of visual data easier than ever. Such pervasive exchange of visual data, however, has increasingly raised questions on how sensitive visual information can be protected. In this project, we aim at developing a computing framework so that distrusted parties on a network can keep their imagery data private but at the same time collaborate with each other on a distributed image processing task.

Secure Image Processing

The basic approach is to design clever Disguising Transformations P(). If Alice has private content I and Bob has proprietary algorithm f() with private B, the disguising transformation P() allows Bob and Alice to exchange information to compute f(I,B) without disclosing the true nature of the data. Such computation is called Secure Multiparty Computation in cryptography and is usually very computationally intensive. Our innovation in this project is to devise new efficient algorithms that can be used for real-time signal and image processing.

Active Research Areas:

Relevant Publications

Collaborators:

Quasi-Information-Theoretic Security Secure Linear Filtering and Thresholding

Dr. Thinh Nguyen (Oregon State University)