---

In print | April 24, 2008

Mary Prager: When did you find out that you were going to receive this fellowship?

Jeremy Freeman: Well, I got an e-mail from [them] … it’s like this weird thing, they don’t tell you right away, it gets posted on this website before … so one of the grad schools that I applied to, one of the professors … that I had been in contact with, e-mailed me and was like, “congratulations on the NSF.” and then I was officially informed later that night.

MP: When did you apply for it?

JF: Back in November.

MP: And was it a pretty complicated application?

JF: It was fairly complicated. You had to write a series of essays … there was a personal statement, a science-oriented personal statement and a description of proposed activities.

MP: So what was your proposed activity?

JF: I’m very interested in how the brain represents visual information, specifically object information … and a lot of my work has been on the specific case of letters, and how the brain represents the identity of the letter.
So my proposed research was applying a fairly novel FMI imaging data analysis technique to further our understanding of representation.

MP: By novel do you mean new?

JF: It’s a technique that has been developed in the past couple of years and has been applied to a couple of simple cases. The technique is a way to analyze FMI data to decode based on symbols of neuro-activity in people’s brain — meaning, decode which of many letters they were looking at.
So this has been applied to very simple stimuli — simply oriented bars, essentially — so if someone looks at bars that are either horizontal or vertical, it’s kind of like a machine learning an algorithm.

You can use this technique to figure out, based on patterns of neural activity, whether the brain is looking at the vertical bars or horizontal bars. So it’s very neat, very different.
Normally we measure activity in someone’s brain while the person is looking at something, and then try to look at what stimulus evokes what kind of response. With this new technique, we can study a given response and decode what the brain is looking at.

MP: Is this something you want to pursue in grad school?

JF: It’s one of many things I want to pursue in grad school. Fundamentally, I’m interested vision, which works like this: light hits the retina and then the first layer of visual neurons decodes very simple information based on simply oriented bars.
After that, a lot of magic happens and then, we see. Basically, I want to know how that magic happens.

There’s this unbelievably complicated computation that takes the elementary feature information – consisting of edges and motions – and enables us to see it as coherent objects.

Obviously we don’t see a jumble of colors and edges, we see coherent things: chairs and faces and letters. So, I’m fundamentally interested in understanding how the brain puts all that information together to compute the identity of what we’re looking at. I would say it’s currently the most difficult and interesting problem in visual neuroscience.

MP: When did you first get interested in the sciences? Have you always been a science person?

JF: Yeah, since I was fourteen I’ve been interested in trying to understand the nature of perceptual experience. So from a very young age I’ve been fundamentally interested in the sciences from a philosophical perspective, a biological perspective and a neurological perspective.

---

« RETURN TO NEWS

« PREV ARTICLE | NEXT ARTICLE »