Deep learning for pupil and illumination paper published!

October 6, 2020

Congratulations to Kanghyun, Pavi and Colin and our collaborator Ron Appel on completing a nice new paper in Optics Letters!

The team used a deep neural network to jointly optimize the pupil transmission and illumination of a microscope to improve its task-specific performance for detecting malaria within thin blood smears. The optimized illumination and pupil patterns are quite interesting to observe in that their shapes are certainly coupled, and the paper does a nice job at verifying how this coupling matches expectations in simulation. 

Please find more information here: