Research
I work on developing artificially intelligent systems that are able to reason about the visual world. My primary research is in computer vision, touching on human-computer interaction, on fairness/ accountability/ transparency, and recently even on cognitive science. Here are my Google Scholar, CV and a formal bio.
Please visit the Princeton Visual AI Lab page for a list of publications, project descriptions, lab members, and other information. Here's how to get involved in research with us.
Outreach
I spend a lot of time thinking about how we as a society got ourselves into such a diversity crisis in computer science, and how we can get ourselves out of it. Some of my views are briefly summarized in this MIT Technology Review article.
AI4ALL is a nonprofit I co-founded in 2017 to create a diverse future generation of AI leaders. You can read more about the AI4ALL story on our website or in this excellent article in The Atlantic. Here's how to get involved with AI4ALL.
My students and I are also involved in a number of community initiatives to provide pathways into AI.
Random Tidbits
One of the most useful books I ever read is Stress-Free For Good by Fred Luskin and Kenneth R. Pelletier. It makes a really solid point that while mental stress may be helpful for motivation, physical stress (heart pounding, muscles tightening, sinking feeling in your stomach) is strictly counter-productive on every level — except when you're running from an actual physical tiger, which you probably aren't. So the book describes some straight-forward techniques to trick your body into being less physically stressed.
A 2015 study published in Science and extensively covered by the media led by Sarah-Jane Leslie found that scientific fields where innate brilliance is believed to be required tend to attract fewer women and racial minorities. This is an incredibly important but deeply distressing finding. It's particularly frustrating since innate brilliance is not even really a thing. For example, check out Peak: Secrets from the New Science of Expertise by Anders Ericsson which essentially invalidates this concept entirely. It summarizes several decades of research into how the right type of practice can be used to develop almost any skill.
There are some excellent resources which seek to illuminate both the hard-hitting impact and the deep structural causes of AI bias. In particular, I highly recommend two great books, Race After Technology by Dr. Ruha Benjamin and Algorithms of Oppression by Dr. Safiya Umoja Noble. If you only have a few minutes, please watch AI, Ain't I a Woman by Joy Buolamwini for a beautiful and painful look at the topic. This is why it is so vitally important that we create space for a diverse next generation of AI leaders.
Say Hi
Office: CS 408 at 35 Olden St
Email: olgarus at princeton dot edu
Please be sure to read this about research at Princeton Visual AI before emailing me.