Face of the Team – Diversity, Equity, and Inclusion
Tsz Hin Martin Cheung, Erik Noyes, Leonidas Deligiannidis · 2021 International Conference on Computational Science and Computational Intelligence (CSCI) · 2021
Among the many breakthroughs in machine learning, and specifically generative deep learning, trained neural networks can render entirely synthetic, photorealistic faces -- reflecting a deep understanding of face compositions and variability. This deep understanding and rendering power can be harnessed for powerful social commentary about Diversity, Equity, and Inclusion (DEI) in the business landscape, and support A.I.-enabled activism about inequalities in corporate leadership, in venture finance, and the overall success and failure of DEI initiatives. Properly trained generative deep learning models can enable the interpolation of latent space/latent dimensions in facial structures, thus enabling the hybridization of management team facial attributes to depict team diversity in an "averaged" team face. This project leverages state of the art, pre-trained generative face-creation models to support multi-variable interpolation between submitted faces, e.g., gender, race, age. The contribution is to show a powerful new way to track the (non)diversity of powerful management teams.