Leyla Kabuli
University of California, Berkeley
lakabuli@berkeley.edu
Bio
Leyla Kabuli is a fifth year PhD candidate in EECS at UC Berkeley advised by Professor Laura Waller. She received a BS in EECS and a BA in Music from UC Berkeley in 2021, graduating as the 150th University Medalist. Her research is in computational imaging system design, at the intersection of information theory, machine learning, and optics. She is supported by the National Science Foundation Graduate Research Fellowship Program and the Berkeley Fellowship.
Areas of Research
- Signal Processing
Designing imaging systems to maximize information capture
Computational imaging systems combine optical hardware with computational processing, pushing the limits of image capture and interpretation by transforming raw, often non-human-interpretable measurements with powerful reconstruction algorithms. My research focuses on developing fundamental and interpretable methods for computational imaging system design. Using information-theoretic techniques, I study applications ranging from biomedical imaging to radio astronomy, with the goal of optimizing optical hardware to capture the best measurements for downstream processing and inference. This work focuses on our data-driven mutual information estimation framework, which bridges a crucial gap in imaging system design by enabling direct evaluation of raw measurements without requiring computationally-intensive reconstruction algorithms or ground truth data. Using mutual information estimation, I uncover tradeoffs and design principles for computational imaging systems and present methods for automatic optical element optimization through differentiable and black-box approaches.