Researchers at Guelph University and the University of Toronto Mississauga in Canada have built a neural network that mimics the fruit fly's visual system, and can distinguish and re-identify flies.
The new network could enable thousands of laboratories that use fruit flies as a model organism to conduct more longitudinal work.
The researchers built their algorithm by combining insect biology and machine learning to process low-resolution videos of fruit flies over two days.
The algorithm reliably identified the same fly on the third day with an F1 score of 0.75, which is only slightly lower than scores for algorithms without the limitations of fly-brain biology, while outperforming human biologists.
Toronto's Joel Levine said, "The approach of pairing deep learning models with nervous systems is incredibly rich. It can tell us about the models, about how neurons communicate with each other, and it can tell us about the whole animal."
From Canadian Institute for Advanced Research
View Full Article
Abstracts Copyright © 2018 Information Inc., Bethesda, Maryland, USA