Thanks to recent advances in machine learning, facial-recognition technology can pick faces out of a crowd with impressive accuracy. But just how far away from their subjects can cutting-edge facial-recognition systems work? According to a new report, a whole lot further than you probably think. So far, in fact, that the person identified may not even realize that they were caught on camera in the first place.
New Scientist writes that the United States military is in the process of funding the creation of a portable face-recognition device that’s able to identify individuals from up to 1 kilometer away. That’s the equivalent of almost 11 football fields.
The project, titled Advanced Tactical Facial Recognition at a Distance Technology, began in 2016. A working prototype of the system was demonstrated at the end of last year. Research by the U.S. Special Operations Command (SOCOM) is reportedly ongoing, although there’s no word on when this technology will see the light of day.
The tech has been developed by Arlington, Virginia-based firm Secure Planet. On its website, the company states that its “mobile biometric verification and identification system provides critical protection to soldiers and civilians, allowing them to stay safe in ever-evolving high threat environments.” It adds that it is “committed to providing innovative products that will forge our way to a more secure planet.”
The firm doesn’t reveal exactly how its technology works. Even with cutting-edge A.I., it’s a massively challenging problem to solve, since adding longer lenses to cameras increases noise from the vibrations. At these kinds of distances, atmospheric turbulence can also pose a problem when it comes to distorting the image in an effect similar to the heat shimmer seen on hot days. To get around this, a separate neural network could potentially be used to unscramble the distorted image in order to get workable facial data for identification purposes.
It’s also not clear exactly how the technology will be deployed. While it was originally designed for handheld use, it could also reportedly be used with drones. SOCOM documents say that it could possibly be shared with law enforcement agencies.
San Francisco startup Dimensional advertised for people to walk robot dogs through city streets. The New York Times reported that the job could include carrying a roughly 33-pound robot back to the office if it overheated. A lift home is an awkward requirement for a labor-saving machine.
Keeping the hardware running is only one challenge. Anthropic's robotics tests found that language models mostly failed when controlling joints directly, but did better when directing trained controllers. Giving an AI agent a body doesn't teach it to handle ordinary household objects.
Soon, a mattress protector may be more than a bed cover. Australian medtech company Sleeptite and researchers at RMIT University have received a US patent for FlexiSense, a smart bedding technology that tracks body position and posture to help caregivers identify when older residents may be at risk of falling.
The technology is being developed for Sleeptite’s in-bed monitoring system, REMi, and was recently on display at The MedTech Conference in Boston. A peer-reviewed study in Flexible and Printed Electronics says the sensors can also be made at scale without sacrificing performance or durability.
AI companies spend enormous amounts of time convincing us their technology is safe and secure. But what happens when those same companies become targets of increasingly sophisticated AI-powered attacks? Anthropic, the company behind Claude, is now facing that question. According to a Washington Post report, cybersecurity startup Cathedral discovered vulnerabilities that let its researchers access sensitive information in Anthropic's internal systems.
The researchers used AI agents to identify potential entry points in less than 12 hours. Interestingly, they didn't need to crack Claude itself. Instead, they exploited weaknesses in third-party business applications that Anthropic uses for everyday operations. Anthropic's security testing program disclosed the findings, which allows outside researchers to identify and report vulnerabilities.