D3.2 Early Fusion and Mining v1

This deliverable contains the first version of algorithms for early fusion and mining of multi-modal sensor data, in order to extract and recognise specific patterns and detect unusual and critical events. Initially, work is described to combine the Philips DTI-2 data and the Gear4 Renew Sleepclock in order tomeasure parameters of the PwD that relate to their daily activity patterns. Work on WIMU sensors is presented, describing algorithms created for processing accelerometer, gyroscope and magnetometer data. Finally the CEP engine is introduced. CEP is used to combine information and infer events. Detailed implementation and results are shown and future work is discussed.

© 2011 Dem@Care | Centre for Research and Technology Hellas :: Information Technologies Institute