Connected Media @ SRI

 
Connected-Media-at-SRI-small_03 Cloud computing and mobile devices will combine to create a cloud media platform that will offer the on-demand consumption of rich and immersive media across many end-user devices (phone, tablet, TV, automotive). Improvements in networking technologies will enable the delivery of high resolution and multi-view video in real-time, dynamically adapted to changing network conditions and multiple screen types. New modalities of interaction such as natural user interfaces enabled by Kinect like technologies promise a leap forward in ease of interaction across a diverse range of applications from health to retail. Advances in analytics will enable digital systems to monitor, analyse and thereby react in a more intelligent fashion to the needs of users.
Research Areas
To meet the complex set of research challenges thrown up by these trends SRI conducts research in the following areas
Cloud Media Platforms
This research area investigates approaches to ensure media delivery with high QoE with a particular emphasis on cloud based media delivery. Specific topics include media transmission and adaptation mechanisms and protocols, cloud architectures to ensure optimized mobile media delivery, cloud media processing to augment media delivery, collaborative platforms for immersive media, multi-media service composition etc.
Intelligent Interactivity
This research area investigates how emerging interaction technologies can be applied to develop smarter forms of human-technology interaction in application areas. Specific topics include the use of natural user interfaces technologies and augmented reality (AR) to improve user interfaces to digital systems (in health, e-learning, TV etc), the use of data stream mining to adapt system response to users etc.
Cross Platform Media Application
The multiplicity of platforms and arising device heterogeneity means that app(lication) creators have to develop many variants i.e. “cross platform”, if they want to address all markets. The problem is not confined to smart phones but includes tablets, games consoles, set top boxes and, recently, connected TV. This theme researches the use of web/browser based development environments using HTML5 media features as a solution to the cross platform dilemma.

3DSensing
3D Sensing entails the use of 3D natural user interface techniques (multi-touch, gesture), such as those based on Kinect, with computer vision, speech recognition and other sensing modalities (location, orientation) to recognize and track human interactions thereby creating many new possibilities for innovation across a range of fields. COMAND is actively researching novel uses of 3D sensing in combination with cross platform development techniques in the retail, sports and e-Health fields using both proprietary (Kinect) as well as cross platform open source (OpenNI) technologies

Interoperability for Internet of Things
Interoperability of Internet of Things investigates solutions to enable communications and co-operations between IoT systems with different implementation language, execution environment, or model abstractions. Interoperability in the Internet of Things brings many challenges due to the multiple protocols and standards. Against this background industries in all types of domain from health to smart grid are struggling for communication solutions to leverage the opportunities provided by the Internet of Things. COMAND focuses on the development of multi-application multiprotocol Internet gateways and control systems.

 
Projects
SVDN – A platform for scalable video distribution over P2P networks. Funded by Enterprise Ireland.

MVDN – A platform for multiview video over P2P networks. PhD extension of the SVDN platform.