PamiTC Job Board - Posting Details

Title: 3D Activity Modeling / Machine Learning for Smart HospitalsPosted: January 15, 2014
Company/Institution: Johns Hopkins University

Description: Two multi-year post-doctoral or research scientist position in 3D Activity Recognition for Smart Hospitals is available at Johns Hopkins University. This is a large and unique multi-faculty collaboration between faculty in Computer Science, Public Health and Medicine.

The goal of this funded project is to develop methods to reduce patient safety errors in the hospital. This environment is monitored using an array of sensors including the Kinect. Research problems of interest include single- and multi- sensor robust activity recognition, modeling inter-user variability, robust tracking in crowded environments and learning domain specific priors for scene understanding.

REQUIREMENTS: Candidates must have a PhD in computer science, electrical engineering, statistics, biomedical engineering or related disciplines. The ideal candidate for this position should have a) a strong background in low-level image processing and should have published work in top computer vision venues like CVPR and ICCV, b) interest in software engineering as the project will require practical demonstrations of the proposed research, and c) strong communication skills as the candidate will get the opportunity to learn and engage with a multidisciplinary team of faculty colleagues. Exposure to human activity modeling or machine learning (probabilistic modeling approaches, structured prediction, dimensionality reduction) would be desirable. Graduates with an interest in moving towards healthcare applications are especially encouraged to apply.

RESOURCES: The candidate will have the chance to work closely with professors in machine learning (Prof. Suchi Saria), computer vision (Profs. Rene Vidal and Greg Hager) and medicine (Prof. Peter Pronovost, MacArthur fellow). This project will take place in the Laboratory for Computational Sensing and Robotics (LCSR). LCSR is home to 80 faculty, students, postdocs, and staff, and is known as one of the preeminent institutions worldwide integrating engineering and medicine. In addition, you will also have the chance to interact with faculty in Machine Learning and in the Armstrong Patient Safety Institute.

This project will expose candidates to the emerging and growing application area of computational healthcare and medicine. They will also acquire the necessary skills to lead multidisciplinary research, valuable in seeking positions downstream within academia and industry. The chosen candidate will also be encouraged to seek out new research opportunities within LCSR and to develop independent research projects. Hopkins Engineering is located a short ten minute drive from the Baltimore waterfront with a number of activities geared towards students and young professionals. Student housing is also available in Charles Village, right next to campus.

Application Instructions: Interested candidates should send a copy of your CV, transcript and 1-2 letters of reference to Prof. Saria at ssaria AT