The Johns Hopkins University and the Applied Physics Lab are seeking highly creative and motivated Postdoctoral Research Associates to develop, implement, and evaluate a novel mathematical framework, based on probabilistic graphical models, for the interpretation of multiple data streams in complex, noisy and partially observable environments. The primary research areas for the positions are:

• Probabilistic graphical models that integrate multiple sensor modalities, accommodate model learning with prior knowledge, and support a rich query language
• Probabilistic graphical models for categorization of objects in images
• Bayesian decision theory for sensor planning and mission- sensitive allocation of sensor resources.
• Software prototyping of parallel and incremental inference algorithms

This position will be with Profs. René Vidal (BME), Andreas Terzis (CS), Yanif Ahmad (CS), and Dr. I-Jeng Wang (APL) at JHU. The work will be done in collaboration with Profs. Stuart Russell and Shankar Sastry at UC Berkeley and Dr. Avi Pfeffer at Charles RIver Analytics. The successful candidate should have a Ph.D. in a relevant quantitative field (e.g. Computer Science, Applied Math and Statistics, Electrical Engineering), a strong publication record in machine learning and Bayesian statistics, and strong interests in systems prototyping and applying these techniques to real-world datasets. A background in discrete optimization, Graph Cuts, MCMC methods, POMDPs, inference algorithms and nonparametric Bayesian methods is also desirable. The appointment will be for one year initially (renewable) starting September 2011 or as soon as possible thereafter.

Applications must be sent by email to Yanif Ahmad <yanif@cs.jhu.edu>, and must include a brief letter of application, a CV including a list of publications, and names and email addresses of 2-3 referees. Review of applications will begin on July 15th, 2011.

Note: If you are attending CVPR 2011 conferences please feel free to contact Prof. Rene Vidal about these positions.