I am Devipriya Raju, a thesis-based Master’s student in the Department of Computer Science at the University of Alberta.My research interests include computer vision, multimodal representation learning, and applied deep learning for modeling complex visual and time-dependent systems.
My current work focuses on the temporal prediction of Multiple Sclerosis lesion growth using boundary-conditioned variational autoencoders with geometry-semantic attention mechanisms.This approach aims to help clinicians forecast disease progression up to 12 months in advance, enabling more proactive and personalized treatment planning.This formulation treats disease progression as a conditional generative sequence modeling problem, combining representation learning with clinically meaningful uncertainty estimation.
Outside academia, I work in animal rescue and rehabilitation, training abused stray dogs to become adoptable.