DeepLabCut: markerless pose estimation of user-defined body parts with deep learning

Alexander Mathis, Pranav Mamidanna, Kevin M. Cury, Taiga Abe, Venkatesh N. Murthy, Mackenzie W. Mathis & Matthias Bethge

Nature Neuroscience (2018), doi.org/10.1038/s41593-018-0209-y

Central to my approach to studying Drosophila behavior was in the quantitative analysis of behavioral kinematics. To achieve this, I established a collaboration with colleagues at Harvard University to develop DeepLabCut, a deep-learning algorithm for the automated tracking of body posture from videography. This innovative tool has been a game changer for the study of behavior and has become a worldwide standard in the field, with this original publication amassing over 2,700 citations.