Persistence and Perseverance for a PhD

May 25, 2011

Nadia Payet Nadia Payet has had a long journey — literally and figuratively — to graduation.

Payet is a native of Réunion, a tiny French island in the Indian Ocean, more than 10,000 miles and two days of travel away from Corvallis. And her journey to graduation has taken a few twists and turns along the way.

Payet was lucky enough to travel to several countries before she even finished high school: France, Greece, Germany, Australia, Canada and the east coast of the United States. So the decision to attend school in Oregon was not at all daunting.

She arrived at Oregon State University in 2004, as a graduate student in computer science on an exchange program. Originally slated to be in Corvallis for just six months, Payet fell in love with Oregon and decided to continue her studies at OSU to obtain a PhD in computer vision.

And then the bottom fell out from under her.

Three years into her program, Payet’s major professor left OSU and she was unable to continue her research with him. And although the other computer science professors at OSU were very supportive of her, their research interests and projects didn’t match up with the work she had been doing, so none of them were able to take her on as a graduate student.

Reunion Island Réunion is a tropical island in the Indian Ocean, east of Madagascar. It is one of the 27 regions of France.

“I thought about just doing enough work to get a master’s degree and just go home,” recalls Payet, “but I really wanted to get a PhD because I love research, and that’s what I want to do.”

Encouraged by friends, family and the computer science faculty, Payet continued on at OSU in a holding pattern for the next year and worked with professor Tom Dietterich on one of his research projects.

In 2008, the School of EECS hired assistant professor Sinisa Todorovic, who was interested in mentoring Payet, but not with the research she had done before. “I had to scrap all of my previous work and start all over again,” says Payet. “But getting a fresh start gave me a new boost and I was super pumped to get going.”

Todorovic says that Payet’s perseverance showed great strength of character. “I didn’t feel any resentment from Nadia about having to start from scratch,” he says. “She has always been very optimistic and has a great attitude.”

Payet’s research specializes in object recognition — something that is easy for people to do, but not so for a computer. Looking at a photo, a person can easily identify a chair or a desk. Object recognition gets computers to “localize” these objects, label them, and “understand” what it looks like in 3D (i.e. that the side view looks different from the front view).

Traditionally, object recognition has been done by having computers analyze the colors in an image: a basketball, for example, is easy to recognize by its color. Payet’s research gets computers to look at the shape of objects.

“The contour of a car tells you that it’s a car; color isn’t important in that situation,” explains Todorovic. “Nadia’s research helped us show that if you use shape as well as color in recognizing objects, the results are much better.”

Object recognition Nadia Payet's research helps computers identify objects in images.

“Object recognition tells the basic story of what’s in a picture,” says Payet. “These basics can then be used to build better things.”

Many cameras already have face recognition capabilities built into them so that the image focuses on people, rather than something in the background. In the near future, more detailed face recognition capabilities could be used, for example, at airports to identify individuals.

In each of the three years she has worked under Todorovic, Payet’s research has been recognized by the computer vision research community with many publications in famous conferences. Payet was also recently chosen as one of approximately 20 students nationwide to present her work to senior members of the research community at the “Doctoral Consortium” of the IEEE Conference on Computer Vision and Pattern Recognition.

Her journey at OSU has taken seven years, but Payet says it has all been worth it and she is grateful for the people at OSU and in the community. She will take the lessons she has learned in the classroom — and in life — to her job at Amazon in Seattle, to work in a research lab on the Kindle.

But she’d like to build her own company in Réunion someday and be able to contribute to the community — a place she calls paradise. “If people go away to learn new things and nothing ever comes back, it’s a waste,” Payet says. “But first the geeky part of me wants to do really technical stuff.”


— By Gale Sumida