We are looking for 20000 hi-resolution images to be sourced from the US and annotated.
The images to be equally divided as per diversity criteria
Faces dataset
Quantity: 20,000 images
Format: high(ish)-resolution photos, 1000x1000p minimum
Factors of diversity to balance:
Race/Ethnicity
Color of skin
Facial features
Hair texture
Style
makeup/no makeup
piercings/no piercings
Hats, turbans
Hair colors and styles (eg dreadlocks, non-natural colored, mohawks, braids, afros, etc)
Gender presentation (masculine, feminine, agender, androgynous, bigender,)
Beards, mustaches vs. no facial hair
Bindis
Tattoos
Glasses (sun or prescription)
Visible Disabilities/Abnormalities
Downs syndrome
Cochlear implants
Cleft palates
Vitiligo
Self-Reported Metadata
Gender (select one or more):
Woman
Man
Non-Binary
Transgender
Hijra
Two-Spirit
Other (please specify)
Race/Ethnicity (select one or more)
Black American
Black Non-American
Indigenous
East Asian
Latin(a | o | x)
Middle Eastern
Pacific Islander
South Asian
White
Prefer not to say
Other (please specify)