Face.com Aims To Make Fake IDs Useless With Age Detection in API
Face.com is adding age detection to its API. This news was delivered to us in an email with the subject, “Fake ID’s useless thanks to Face.com (college Freshmen mourn).”
“Yep, no more lying about your age,” a spokesperson tells us.
“Age Estimation was added to provide an estimation representing an age group for a detected face,” Face.com CEO Gil Hirsch tells WebProNews. “The new Face.com API provides a minimum & maximum ages, an estimated age, and most importantly a confidence level on the range of 0%-100% of how accurate this result is (derived from various quality parameters like lighting and pose).”
“Our intent is for developers to decide for themselves which confidence levels to use as a threshold, and in any case refer to the max & min estimates for the approximate age group,” he says.
We asked Hirsch about the chances of a developer wanting to use the API for age verification getting inaccurate results, which could result in potential legal trouble, and if there is a policy in place to prevent developers from using it for such purposes.
“One prominent use-case our client services were trying to achieve with age estimation was to provide increased moderation and safety features such as kids appearing in video chat and photos shared by users,” he tells us. “It is up to these services to decide how such results will be used, if at all. In general, Face.com focuses on commercial uses that work well on mass amounts of image data, and our focus is to provide the best estimated results along a disclaimer that such results may not be accurate.”
When asked about what ways developers are prohibited from using the API, Hirsch says, “The API is restrictive in a number of ways. In short we do not provide the means to scan the entire internet or identify a stranger on the street through these basic limitations: (1) Limit of images that can be processed, (2) Limit recognition to ‘people you know’ (friends on Facebook, a personal contact list, etc).”
Face.com, based in Palo Alto, launched in 2009. It has 10 employees, and has over 40,000 developers using its API. It also powers the Photo Finder and Photo Tagger Facebook apps.