Private image computation: The case of cloud based privacy-preserving SIFT
Zhan Qin, Jingbo Yan, Kui Ren · 2014
In this paper, we present SecSIFT, a high-performance cloud based image feature detection system for performing Scalar Invariant Feature Transform (SIFT) over private image data without compromising the privacy. In contrast to previous works, we outsource the computation of image feature detection to a set of independent, co-operative cloud servers, and keep the outsourced computation procedures as simple as possible. Using this framework, we are not restricted by efficiency limitations of homomorphic encryption scheme and thus are able to implement applications such as social discovery or behavior prediction with less complexity on computation and communication.