MSc (Tech) Vivienne Huiling Wang has been developing methods for tracking and segmenting objects in videos since her MSc thesis. This is one of the most challenging topics in computer vision, but she has combined combinatorial optimization and deep learning effectively.
During her PhD thesis project, Wang got interested in another emerging field of AI – robot learning. A major challenge in robot learning is that there is much less available data for robots than for computer vision, audio and speech, and natural language processing. The reason is that for audio, images, and language the internet is full of data that can be used. However, the amount of data for robots is limited, and it is slow, expensive, and dangerous to collect more.
Wang investigated how the limited amount of data could be used creatively so that the robots could learn from a small number of examples. She has developed such methods, and her work is a part of the new wave of reinforcement learning for robotics.
Wang’s PhD thesis paves the road for new scientific ideas and practical breakthroughs.
“Research on robotics and computer vision is my dream job. There are still many ideas and breakthroughs to make,” says Wang.
Vivienne Huiling Wang continues her research on robot learning at Aalto University.
Public defence on Friday 28 March
The doctoral dissertation of MSc (Tech) Vivienne Huiling Wang in the field of signal processing titled Robust Visual Perception and Decision Making for Autonomous Systems will be publicly examined at the Faculty of Information Technology and Communication Sciences at Tampere University on Friday 28 March at 12.00 at Hervanta Campus, Rakennustalo, auditorium RG202 (Korkeakoulunkatu 5, Tampere).
The Opponents will be professors Melih Kandemir (University of Southern Denmark) and Luka Cehovin (University of Ljubljana, Slovenia). The Custos will be Professor Joni Kämäräinen from Signal Processing Research Center.