DeepLabCut AI Residents for Next-Gen Animal Behavior
Mackenzie Mathis & Alexander Mathis (École Polytechnique Fédérale de Lausanne (EPFL))
To develop a DeepLabCut AI Residency Program for underrepresented groups in machine learning and computer science in order to recruit, fund, and nurture the next generation of open source leaders.
DeepLabCut is a deep-learning based software package for analyzing animal behavior with state-of-the-art machine vision. DeepLabCut is used for research in a growing range of fields: from animal conservation, to biomechanics, to neuroscience. Many end users do not have formal training in computer science, deep learning, or experience with best practices in programming. The team currently receives dozens of spontaneous applications per year asking to spend time in their laboratory to learn the principles of the tool and/or to help develop and maintain the code base. This proposal aims to develop a formal program, the DeepLabCut AI Residency Program, to host paid interns exclusively selected from groups underrepresented in computer science, such as women, those with non-traditional backgrounds, diverse ages (career stages), gender identity, ethnicity, religion, disability, and sexual orientation, among others. The goal of this DeepLabCut AI Residency will be to train and empower users to use and modify code, develop new add-ons, and become leaders within the DeepLabCut consortium and the open source community more broadly. To this end, the Residency Program will aim to recruit both more experienced users and novice users to provide a nurturing environment where all can learn more about DeepLabCut, open source practices, and develop new skills.