Distributed MBARK Based Mobile Face Recognition

Perform a feasibility study on developing a face recognition application in a cloud computing environment utilizing MBARK as the middleware. The application will have a client piece that will run on a mobile device and a server piece that will run on a higher-performance computing cluster. The specific objectives of the feasibility can be enumerated as follows: 1. Split/distribute the MBARK middleware to allow local functions – such as sensor data capture and filtering – to pass data and control streams with server functions such as compute-intensive analysis and database matching. 2. Create a face recognition workflow in the XML format as required by MBARK. DSCI’s STARFace SDK will be used to perform the face recognition functions on both the client and server sides. 3. Define additional UI component(s) that would be added to the MBARK UI to show image snapshots and matching results.