Institutional Repository [SANDBOX]
Technical University of Crete
EN  |  EL

Search

Browse

My Space

Personalized motion sensor driven gesture recognition in the FIWARE cloud platform

Petrakis Evripidis, Sotiriadis Stelios, Kostas Stravoskoufos, Alexandros Preventis

Full record


URI: http://purl.tuc.gr/dl/dias/7F7F64BE-4B88-4D26-8F5D-93AF03A72C23
Year 2015
Type of Item Conference Paper Abstract
License
Details
Bibliographic Citation Alexandros Preventis, Kostas Stravoskoufos, Stelios Sotiriadis, Euripides G.M. Petrakis.(2015, Jun.). Personalized Motion Sensor Driven Gesture Recognition in the FIWARE Cloud Platform. Presented at 14th International Symposium on Parallel and Distributed Computing (ISPDC 2015). [Online]. Available:http://www.intelligence.tuc.gr/~petrakis/publications/ISPCD2015.pdf
Appears in Collections

Summary

Gesture recognition technology enables new means of user communication and interaction with machines. This work focuses on gesture recognition by analyzing data, obtained by motion sensors, in the cloud. We present Interact, a cloudbased gesture recognition system that uses FIWARE. To promote the development of Future Internet (FI) applications, all functionalities of Interact are offered as cloud services, enabling elasticity, lower cost of maintenance and off-site efficient datastorage. To promote productivity, Interact offers a REST API for recognizing, storing and managing customized gesture collectionsin the cloud as well as for subscribing to sensors. The system is sensor independent and is designed to be compliant with the most popular motion sensors (i.e., Leap Motion or Kinect). To demonstrate the functionalities of Interact, a system prototype has been developed utilizing the LEAP Motion sensor, offering all the previously described functionalities. The prototype is hosted on FIWARE Lab and is available for testing.

Services

Statistics