Violetta Shevchenko
![Violetta Schevchenko](/aiml/sites/default/files/styles/ua_image/public/media/images/2019-04/190328_AUNI_AIML_053.jpg?itok=5BxDdA7E)
PhD student
What are you working on?
In my research I’m tackling the challenge of visual question answering (VQA)—a key test for artificial intelligence (AI). If we get VQA right, we’ll be able to show AI any image, then ask it a question about what’s happening in that image—in normal, everyday language—and the AI will be able to answer it.
Current methods are quite restricted by the range of images and questions that they can understand. I want to improve VQA algorithms so that they can constantly learn, incorporate new data and use additional information, making them more robust and applicable for real-world problems.
VQA systems have diverse practical applications, the main purpose of which is to assist people. This can range from applications that help visually impaired people perceive and utilise visual data, to systems that simplify human-computer interaction.
How did you come to be here?
After completing a masters in computer science at Southern Federal University in Russia I was keen to do a PhD focused on computer vision and machine learning at a world-recognised research institute. I liked the look of the University of Adelaide’s Australian Institute for Machine Learning (AIML)—then known as the Australian Centre for Visual Technologies—so contacted the institute’s director, Professor Anton van den Hengel, and arranged a visit.
I came and had a look at the University and AIML’s research and talked to lots of students. I was very impressed with what I saw and heard, now here I am!
What do you like about AIML?
I like the AIML research environment; both the facilities and the people I work with. We have various meetings, workshops and social events where we share ideas, discuss progress and look for new collaborations. We have experts in many different computer science areas, which means I can always find someone to talk to about my work and ask for advice, no matter what problem I’m facing. It makes me feel part of a big, friendly research team and helps with my work.