suzie petryk

/ Github / LinkedIn / CV /
spetryk at berkeley dot edu

Hello! I am a Ph.D. student in AI at UC Berkeley, co-advised by Joseph Gonzalez (in machine learning systems) and by Trevor Darrell (in computer vision). I'm fortunate to be part of the RISE lab (Real-Time Intelligent Secure Explainable Systems) as well as BAIR (Berkeley Artificial Intelligence Research). Before coming to Berkeley, I obtained my B.S. in computer science at Cornell University.

My research interests are currently in using explainability of convolutional neural networks for model compression and faster training. I am also interested in neural architecture search.

Outside of research, I love to climb. I also like going for a run when it's nice weather (very grateful for Berkeley's average of 256 sunny days per year).

research

I started my research career in materials science before switching focus to machine learning and computer vision. I spent a summer at the University of Utah growing crystals of a special conductive material (the best part: sealing vacuum-pumped glass tubes with a blowtorch!). I then spent a summer at UC Berkeley in transportation engineering. I started as a PhD student at UC Berkeley in August 2019 and am excited about my current research direction on improving the efficiency of CNNs.

publications

presentations

  • Lane Blockage Detection Using Loop Detector Data
    Ivy League Undergraduate Research Symposium, 2018. Oral Presentation.
  • Growth of Single Crystal Topological Insulators and Dirac Semimetals
    National Conference on Undergraduate Research, 2017. Poster Presentation. Sponsored as winner of poster competition in materials science summer research program at University of Utah.

teaching