Google's smile-detection model used a racial classifier trained on four race subgroups and a gender classifier with two categories.
Notes on verification
Directly confirmed by original Google research paper (arXiv:1712.00193) and corroborated by multiple independent secondary sources (The Register, Springer Nature, Techmonitor). [tier=gold indep_score=0.983 clusters=3 claim_tier=notable]
Sources
- Google’s New AI Smile Detector Shows How Embracing Race and Gender Can Reduce Bias (seed:technology_and_ai)
- https://www.theregister.com/software/2017/12/06/google-learns-to-smile-because-ais-bad-at-it/798222 (corroboration)
- https://arxiv.org/pdf/1712.00193 (corroboration)
- https://ar5iv.labs.arxiv.org/html/1712.00193 (corroboration)
- https://link.springer.com/chapter/10.1007/978-3-030-11009-3_35 (corroboration)