Massive databases containing tens of thousands of photos from public security cameras, OKCupid and other sites used to train facial recognition without users’ knowledge, report claims
- Databases used to train facial recognition take people’s faces without consent
- One data set, Brainwash, used 10,000 images taken from a coffee shop’s camera
- Other dating sites like OKCupid have also been the target of researchers
- Some companies have begun to rethink their use of pictures without consent
Published: 17:13 BST, 15 July 2019 | Updated: 17:16 BST, 15 July 2019
With images aggregated from social media platforms, dating sites, or even CCTV footage of a trip to the local coffee shop, companies could be using your face to train a sophisticated facial recognition software.
As reported by the New York Times, among the sometimes massive data sets that researchers use to teach artificially intelligent software to recognize faces is a database collected by Stanford researchers called Brainwash.
More than 10,000 images of customers at a cafe in San Francisco were collected in 2014 without their knowledge.
OKCupid and photo-sharing platforms like Flickr are among for researchers looking to load their databases up with images that help train facial recognition software. Stock image
That same database was then made available to other academics, including some in China at the National University of Defense Technology.
Those researchers have helped to develop the country’s AI-based surveillance software used in China’s efforts to track one of the country’s ethnic minorities.
While Brainwash was recently taken down by researchers, the practice of harvesting users’ faces without their consent continues.
The Times reports that users of the dating site, OKCupid had their pictures co-opted to help create an AI that can identify the ‘age, sex and race of detected faces’ by a company called Clarifai.
It’s unclear whether OkCupid gave researchers consent to include its users’ images in the database.
As awareness builds around the use of unwitting peoples’ faces in training facial recognition, some companies have begun to reel in their methods.
Last month, Microsoft has discreetly pulled a facial recognition database from its site that contained 10 million images of some 100,000 people.
The company took down the database after a Financial Times investigation revealed that the database has been used by private companies and military researchers to train facial recognition systems around the world.
The dataset, called ‘MS Celeb,’ included images of ‘celebrities’ pulled from the internet, but also contained photos of ‘arguably private individuals,’ often without their knowledge or consent, the FT found.
Facial recognition is used by governments and law enforcement to help identify people, sometimes matching them against a database of known criminals. File photo
Similarly, IBM was revealed to have roved through Flickr.com — a popular photo-sharing website — scraping user photos to train its own facial recognition software in May. None of those users were notified that they had been included in a database.
As a result of increasing sophistication and awareness of facial recognition, its deployment, particularly by law enforcement and other public institutions, has recently risen as a political issue.
Earlier this year, San Francisco became the first U.S. city to ban its use by police, while members of congress had their first-ever public discussions regarding potential regulations on the technology in May.
HOW DOES FACIAL RECOGNITION TECHNOLOGY WORK?
Facial recognition software works by matching real time images to a previous photograph of a person.
Each face has approximately 80 unique nodal points across the eyes, nose, cheeks and mouth which distinguish one person from another.
A digital video camera measures the distance between various points on the human face, such as the width of the nose, depth of the eye sockets, distance between the eyes and shape of the jawline.
A different smart surveillance system (pictured) can scan 2 billion faces within seconds has been revealed in China. The system connects to millions of CCTV cameras and uses artificial intelligence to pick out targets. The military is working on applying a similar version of this with AI to track people across the country
This produces a unique numerical code that can then be linked with a matching code gleaned from a previous photograph.
A facial recognition system used by officials in China connects to millions of CCTV cameras and uses artificial intelligence to pick out targets.
Experts believe that facial recognition technology will soon overtake fingerprint technology as the most effective way to identify people.