Interactive 3D technologies (I3D) combines research on 3D graphics, computer vision, machine learning, novel hardware, augmented reality and NUI. Our remit is to demonstrate radically new user experiences built using state-of-the-art vision and graphics algorithms, and novel sensing or display hardware. Our projects mix theory and practice, meaning we research new algorithms and techniques, but also build complex systems, and demonstrate our results practically.
“Ultimate” IO This theme explores new designs for cameras for capturing reality, including acquisition of geometry, material properties and light fields, as well as novel display technologies for presenting rich digital content to users.
Reality Remixed This theme tackles the hard technical and user interface challenges in making augmented reality a reality. Key areas are real-time tracking and reconstruction, augmented reality IO hardware, and rich user experiences that blur the physical and digital.
Recovering and Understanding Reality This theme explores new probabilistic models and algorithms to infer physical properties of the scene such as lighting, materials, and semantics such as object recognition and scene understanding. We also explore applications of such methods, particularly for augmented reality.
deForm This theme explores new algorithms for representing and inferring the dynamic nature of the real world, digitally. This includes moving and deforming bodies, hands and faces of users, and our dynamic environments. We also explore new ways in which these dynamic digital representations can be used interactively.
I3D grew out of collaborations between the Sensors and Devices and Computer Vision groups at Microsoft Research Cambridge. We still closely collaborate with these groups, and many of the team hold joint affiliations. We also collaborate closely with: