Microsoft Research PhD Scholarship 2013

In September 2012, Microsoft Research received 111 PhD applications which were reviewed by 153 internal and external reviewers between October and December 2012. We have now selected 20 applications that will be funded through Microsoft Research Connections starting in the academic year 2013/14. Eight proposals relate to the new Joint Initiative with University College London (UCL) and the Joint Initiative in Informatics with Edinburgh University.

The selected applications are listed below.

Expert Visual Classification with Thousands of Categories
University of Bath

Investigation of a Radically New Energy Technology Based Upon Programmable Artificial Photosynthesis
University of Cambridge

Compositional Verification of Scalable Joins by Protocol-Based Refinement
Institute for Software Systems (MPI-SWS

3D Smart Memory for Scale-Out Servers
EPFL

Intuitive and Efficient Design of Enclosures for .NET Gadgeteer
University of Lancaster

Natural User Interfaces for the Developing World
University of Cape Town

Understanding the Dynamics of Embryonic Stem Cells Differentiation: A Combined Experimental and Modeling Approach
University of Cambridge

Exploiting Mobile Sensing and Geo-social information in Mobile Recommendation Systems
University of Cambridge

Future Filesystems: Mechanized Specification, Validation, Implementation and Verification of Filesystems
University of Leicester

All-Pay Auctions in the Real World
Hebrew University

First-Order Satisfiability Modulo Theories
Uppsala University

Learning to Index
Technical University of Denmark

Joint Initiative with University College London

3D Reconstruction of Live Scenes Using Multiple Heterogeneous Mobile Depth Cameras

Probabilistic Databases of Multimodal and Universal Schema

Weakness as a Virtue

Experimental and computational studies on the human antigen-specific T cell repertoire

Joint Initiative in Informatics with Edinburgh University

Performance Portability for Large-Scale Heterogeneous Systems

Provenance for Configuration Language Security

Bayesian Probabilistic Programming for Security

Solving the Problem of Cascading Costs: Better Approximate Bayesian Inference for Data Pipelines