Study indicates neither algorithmic differences nor diverse data sets solve facial recognition bias

Venture Beat | Sep 28, 2020 at 6:50 PM
  • Facial recognition models fail to recognize Black, Middle Eastern, and Latino people more often than those with lighter skin.
  • VGG obtained higher accuracy rates for identifying women excepting Black women and higher rates for men excepting Latino men.
  • After training only on UTKFace, Middle Eastern men obtained the highest accuracy rates followed by Indian, Latino, and white men, while Latino women were identified more accurately than all other women (followed by East Asian and Middle Eastern women).