AWS AI Labs Propose A Method That Predicts Bias In Face Recognition Models Using Unlabeled Data

Algorithmic bias has emerged as a major area of study in artificial intelligence in recent years. An examination of facial recognition software in 2018 defined bias as a disparity in the software’s performance when applied to people of diverse racial or ethnic backgrounds. To check if a face recognition algorithm is biased, it is most […]

The post AWS AI Labs Propose A Method That Predicts Bias In Face Recognition Models Using Unlabeled Data appeared first on MarkTechPost.

ai labs ai paper summary ai shorts amazon applications artificial intelligence aws bias computer vision country data editors pick face face recognition labs machine learning staff tech news technology unicorns usa