Honeywell Labs’ Integrated Security Technology Section is seeking Ph.D. students for two summer internship positions in computer vision and machine learning in our research labs in Minneapolis. A successful applicant will contribute to ongoing work on activity recognition from videos in the wild (e.g. YouTube).  This objective is challenging, in part, because the wide range of activities of interest cannot be addressed with a single feature type or by a single classification technique.  As such, our approach will involve a number of different feature and classification approaches, the results of which must be reconciled to a single classification result.  Applicants with previous experience in score-based fusion of multiple classifiers are strongly encouraged to apply.

Applicants should be currently enrolled in a Ph.D. program concentrating on computer vision and/or machine learning. Understanding of activity recognition literature, past application of data-driven score fusion and possession of strong programming skills in either Matlab or C++ are also desired.  Successful applicants will work in a research environment, exploring novel approaches within the scope of the relevant project.  The methods developed and experimental results obtained over the internship should support a research paper.

Basic Qualifications  

• Ph.D. Student in Engineering Applied Math, or Computer Science, (must be enrolled in a PhD program at a time of internship, no undergraduates)

• Knowledge of Matlab and relevant toolboxes, and/or C++ with OpenCV.

• Research interest in Computer Vision and/or Machine Learning.

• Good written and oral communication skills.

Start of Internship: Exact Date TBD based on candidates availability (May/June 2011)

Duration: 12-15 weeks

Location: Honeywell Labs in Minneapolis, Minnesota, USA

Point of Contact  

Interested applicants should send a CV to Dr. Scott McCloskey (scott.mccloskey@honeywell.com).  Applications will be accepted until the position is filled.