Marina is a Research Scientist working with the Accelerators and VLSI Research Group. Her research focuses on hardware-software co-design and efficient deep learning. This includes designing efficient model architectures, implementing dynamic pruning and adaptive inference techniques, and creating memory and parameter-efficient training methods.
Marina received her Ph.D. and M.S. in Computer Engineering from Brown University. Her Ph.D. dissertation, entitled "AI at the Edge: Efficient Deep Learning for Resource-Constrained Environments", explored innovative techniques for modeling, training, and deploying deep learning models efficiently.
