Researchers have unveiled new Facebook facial recognition technology that will enable people to be recognised even if they are turning their face away. The controversial technology will be able to identify Facebook users based on the clothing they are wearing, their hairstyle and even their body shape.
The technology, which researchers have called Pose Invariant PErson Recognition (PIPER), is thought to be 83 per cent accurate and is able to recognise people without requiring a full-frontal photo. However, researchers have said that the technology is experimental and will not be coming to the social networking website anytime soon.
“This is so far, experimental, long-term research,” said Ari Entin, a representative for Facebook. “I don’t think this is something that we would see any time soon.”
The research may not go down well with some users, and there have already been some issues regarding Facebook’s facial recognition feature. Earlier this year, the social networking website was sued by a user from Chicago, who said that the website had violated people’s rights by using facial recognition without their permission.
However, some experts have praised the research. Anil Jain, a Michigan University professor, said: “It is an excellent piece of research that is pushing the frontiers of face recognition in unconstrained settings.”
“The focus and challenge (both for the government and social media companies) now has shifted towards unconstrained face recognition in unconstrained settings and uncooperative subjects. By this I mean, faces captured from surveillance cameras or photos posted on the social media sites.”
“To work in this space, we cannot rely on traditional face recognition algorithms which relied on ‘face alignment’ based on locations of the two eyes,” he said.
“In unconstrained imagery, eyes may not be visible, e.g., person wearing sunglasses or face is occluded. So, recognition by parts instead of recognition by ‘holistic’ method is needed. This requires that we train a face recognition system with a very large number of images per subject. That explains the approach by Facebook.”
While some may see the research as controversial, facial recognition technology that can recognise people from different angles could be beneficial to law enforcement companies looking for easier ways to identify criminals.