Update: This intern slot has been filled.
This is one in a series of posts advertising internship positions at FXPAL for the summer of 2010. A listing of all internship positions currently posted is available here.
Awareness systems enable people to be aware of others in their sphere. These systems can provide contextual information about the set of people in the system, such as knowledge about someone’s location, their schedule, and what they are doing. Some aspects of a person’s context are more difficult to determine than others. For example, it is relatively easy to determine if a person is using their networked computer or has an appointment on an electronic calendar. However the awareness of what someone is doing is more difficult, often relying on self-report.
As part of research on awareness, we are interested in detecting activities that people do in a public place based on a variety of inputs, including multiple video streams, tracking data, and sensor readings. Activity recognition can aid in awareness of public areas, customization of “smart spaces,” or facilitating activities in augmented places.
We are seeking an intern capable of applying a set of classification methods to detect activities using these inputs, assessing the value of the various inputs for different classifiers, determining the accuracy of the classifications and analyzing the effectiveness of the detection of a range of activities.
Prospective candidates should be enrolled in a PhD program and have significant experience in Artificial Intelligence, Applied Machine Learning, Computer Vision, or Signal Processing with familiarity with probabilistic modeling techniques. Domain knowledge and evaluation skills are a plus.