Nataliya Shapovalova, PhD 2015
“I usually tell people I can’t talk about it,” says Nataliya Shapovalova when asked what she’s working on as a software development engineer at Amazon’s huge Seattle campus. “It’s top secret – but it’s certainly a very interesting new project!”
Despite the cloak of confidentiality, it’s safe to say Shapovalova – who conducted her PhD research in SFU’s Vision and Media Laboratory and whose thesis was titled Towards Action Recognition and Localization in Videos with Weakly Supervised Learning – is generally addressing human-computer interaction and computer vision issues in her career.
“It’s a challenging and complex area. Basically, the research looks at problems of human behaviour and interpretation. And although there’s a large amount of data to work with, we always need more,” she explains, adding that although she had been honing her interest in this area for years, her SFU graduate studies gave her the invaluable opportunity to dive much deeper.
Why did she choose SFU, though? “I was reading a lot of academic papers from there and it just made perfect sense to apply. But I feel I was very fortunate to have been accepted – it turned out to be such a great school for me.”
She attributes much of her positive experience to research supervisor Greg Mori. “He was excellent but he wasn’t the only one. Everyone was very supportive at SFU,” Shapovalova recalls, adding that the algorithm classes she took have proved to be especially useful to her current work.
But one of the most important lessons she learned during her PhD was the value of keeping abreast of wider research topics. “I learned how important it is to read other papers as much as possible. It tells you what other people are doing, of course, but it also helps you understand your own work better.”
Encouraged by her SFU thesis supervisor, she also began moving her academic research into more practical fields, starting with a fruitful five-month internship at Disney Research in Pittsburgh. During her time there, she was strongly focused on looking at how people watch videos.
At Amazon, though, she’s taking this real-world approach even further. “Unlike school, it’s not really an open field of research here. You have to find and refine your tasks a lot more. But I really enjoy this process of digging very deeply into smaller areas, issues and problems.”
And while she describes Amazon as “definitely a cool place to work,” it’s clearly this fertile field of research that’s the main attraction for Shapovalova. “I’m always happy when I’m solving problems and I really want to continue to be part of the computer vision engineering community in the years ahead.”