[ad_1]
Facebook says it will stop using face recognition to tag photos. In a blog post Monday, Meta, the new parent company of the social network, announced that the platform will delete face templates of more than a billion people and turn off its face recognition software, which uses an algorithm to identify people in photos on which they upload Facebook. This decision is a major step in the movement against face recognition, which experts and activists warn has been affected by bias and privacy issues.
But Meta’s message comes with a few big warnings. While Meta says face recognition is not a feature on Instagram and its portal devices, the company’s new commitment does not apply to its meta-universal products, Meta spokesman Jason Gross told Recode. In fact, Meta is already exploring ways to incorporate biometrics into its emerging metaverse business, which aims to build a virtual, Internet-based simulation where people can interact as avatars. Meta also retains DeepFace, an advanced algorithm that powers its photo-tagged face recognition feature.
“We believe that this technology has the potential to enable positive future uses that maintain confidentiality, control and transparency, and this is an approach we will continue to explore as we consider how our future computing platforms and devices can best to serve people’s needs well, “Grosse told Recode. “For any potential future applications of technology like this, we will continue to be public about the purpose, how people can control these systems and their personal data, and how we live up to our responsible innovation framework.”
The fact that face recognition for photo tagging is leaving Facebook, also known as the “big blue app”, is certainly important. Facebook originally launched this tool in 2010 to make its photo tagging feature more popular. The idea was that allowing an algorithm to automatically suggest tagging a specific person in a photo would make it easier than manually tagging and might encourage more people to tag their friends. The software is informed by the photos that people post to themselves, which Facebook uses to create unique face templates associated with their profiles. The DeepFace artificial intelligence technology, which was developed from photos uploaded by Facebook users, helps to match people’s face patterns with the faces in different photos.
Privacy experts expressed concerns immediately after the launch of the feature. Since then, major studies by researchers such as Joy Buolamvini, Timnit Gebru and Deb Raji have also shown that facial recognition can have built-in racial and gender bias and is particularly less accurate for women with darker skin. In response to growing opposition to the technology, Facebook turned on face recognition in 2019. The social media network also agreed to pay $ 650 million last year after a lawsuit alleges that the tagging tool violates the Biometric Privacy Act. information on Illinois.
It is possible that the protection of this particular use of face recognition technology has become too expensive for Facebook and the social network has already received what the tool needs. Meta did not rule out the use of DeepFace in the future, companies, including Google, have already included face recognition in security cameras. Future virtual reality hardware can also collect a lot of biometric data.
“Every time a person interacts with a VR environment like Facebook’s metaverse, they are exposed to the collection of their biometric data,” John Davison, a lawyer with the Electronic Privacy Information Center, told Recode. “Depending on how the system is set up, this data can include eye movements, body tracking, face scanning, voice prints, blood pressure, heart rate, user environment details and much more. It’s a staggering amount of sensitive information in the hands of a company that shows up over and over again, unable to trust our personal information. “
Several of Meta’s current projects show that the company has no plans to stop collecting data on people’s bodies. Meta develops hyperrealistic avatars that people will manage as they travel through the metaverse, which requires tracking the movements of someone’s face in real time so that they can be recreated by their avatar. A new virtual reality headset that Meta plans to release next year will include sensors that track people’s eye and face movements. The company has also weighed the inclusion of facial recognition in its new Ray-Ban smart glasses, which allow the user to record their surroundings while walking, and Reality Labs, Meta’s center for virtual and augmented reality, is conducting ongoing research on biometric data, according to posts on Facebook’s career website.
In addition to Illinois’ biometric privacy law, there are a growing number of proposals at the local and federal levels that could master the way private companies use face recognition. However, it is not clear when regulators will reach a consensus on how to regulate this technology, and Meta will not specify specific legislation that it supports. Meanwhile, the company welcomed the holiday because of its new announcement. After all, this is a convenient opportunity to highlight something other than the recent leak of thousands of internal documents revealing that Facebook is still unable to keep its platform safe.
[ad_2]
Source link