Facial Recognition Software

About

Facial Recognition Software is a type of computer application designed to identify a person by analyzing the subject's facial features within a digital image or video.

History

Starting in 1964, the earliest facial recognition systems were engineered by computer scientists Woody Bledsoe, Helen Chan and Charles Bisson, who created a database of images that associated each person with a list of computed distances between various facial markers. The technology progressed slowly until 1997, when software was developed by computer scientist Christoph von der Malsburg and several graduate students at the University of Southern California and Germany's University of Bochum. The system could recognize faces that were partially blocked by a variet of identification impediments, including facial hair, glasses and sunglasses. In 2006, various facial recognition systems competed against one another at the Face Recognition Grand Challenge.^[2]

Online Presence

Facebook

On June 18th, 2012, the Tel Aviv-based facial recognition company Face.com was acquired by Facebook. On October 28th, 2013, NPR^[1] aired a segment on Facebook's facial recognition systems, which highlighted the various privacy concerns associated with the technology.

Controversies

Nikon Blink Detection

On May 13th, 2009, Taiwanese-American strategy consultant Joz Wang published a blog post^[3] titled "Racist Camera! No, I did not blink… I’m just Asian!" featuring a photo of her Nikon camera mistakenly identifying a blink in her selfie picture (shown below). The photo was subsequently reblogged by Gizmodo^[4] and BoingBoing,^[5] who accused the Japanese camera manufacturer of lacking "tact" and "racial sensitivity."

Google Photos "Gorilla" Label

On June 28th, 2015, Twitter user @jackyalcine^[6] posted a screenshot from his Google Photos app that had incorrectly identified a photo of an African American man and woman as "gorillas" (shown below). In the first week, the tweet gained over 2,800 retweets and 1,400 favorites.

That day, Google spokeswoman Katie Watson released a statement apologizing for the mistake, noting the company had taken steps to remove the possibility of it happening in the future.

"We're appalled and genuinely sorry that this happened. We are taking immediate action to prevent this type of result from appearing. There is still clearly a lot of work to do with automatic image labeling, and we're looking at how we can prevent these types of mistakes from happening in the future."

On June 29th, Google chief architect Yonatan Zunger replied to a barrage of tweets accusing his company of developing racist technology, to which he replied that the software had also commonly mistaken white faces with dogs and seals, adding that "machine learning is hard."

On July 1st, the pop culture site Fusion^[7] published an article by staff writer Charles Pulliam-Moore, who blamed the problem on the lack of racial diversity at companies like Google.

"Perhaps if the titans of Silicon Valley hired more engineers of color, things like this wouldn't happen so often. Or, you know, ever."

Related Memes

NBA 2K15 FaceScan FAIL

NBA 2K15 Facescan FAIL refers to poorly-scanned 3D models of players faces using images captured with the PlayStation Eye or Xbox Kinect in the basketball video game NBA 2K15.

How-Old.net

How-Old.net is a web application developed by Microsoft which uses facial-recognition technology to predict the age and sex of people pictured in photographs submitted to the site. Upon its release in late April 2015, the hashtag #HowOldRobot became a trending topic on Twitter as many users began tweeting about the poor accuracy of the app’s predictions.

Search Interest

External References