Context

The world’s population is scheduled to surpass 8 billion by the year 2030.  There are already that many cell phones and computers around the globe.  My research interests focus on how those devices tell humans apart and how humans use those devices to tell each other apart.  This is the study of biometrics.  Humans perform biometric functions every time they recognize a friend or a colleague, so the idea seems benign to us.  However, teaching machines to accomplish the same task can be a challenge not only from an algorithm standpoint but also from a privacy, legal, and policy point of view.

Additionally, whenever humans and machines autonomously interact with each other, the context of that interaction is hugely important.  To paraphrase Rodney Brooks, the co-founder of iRobot, one of the biggest challenges facing autonomous cars is that they cannot recognize the difference between two teenagers standing on the sidewalk, who might run across the street at any moment, and a mother with her child, who most likely would not.  Obviously, these subtle differences in context have different ideal reactions from the machine.  Computer vision and machine learning are the study of how we can teach machines to recognize that context.

Finally, none of this work would be possible without the ability to recognize and communicate patterns to a larger community.  Statistics and data science enable us to make decisions based on evidence and support those decisions under scrutiny.  Language and visualizations help us make our points to fellow scientists, decision makers, and the public at large.

Videos are a great way to communicate research findings.  Below are a series of videos describing some of my recent research activities.  You can also find a list of my recent publications here.