Top TechTop Tech #106: Thermal Face Recognition, Nokia VR, Cars Sense Pollution
Important innovations in science and technology, every day
By Paul Worthington
Wednesday’s Top Tech:
• Face recognition in thermal images
• Nokia captures VR with spherical camera
• Street View cars sense pollution
Street View cars sense pollution
Those cars Google drives around everywhere aren’t just gathering data for maps and images for the Street View service — they may soon also sense where and when air pollution is at its worst.
Google partnered with sensor network developer Aclima to “enable a paradigm-shift in environmental awareness,” the company says, “by equipping Street View cars with Aclima’s mobile sensing platform to see the air around us in ways never before possible.”
Three Street View cars measured nitrogen dioxide, nitric oxide, ozone, carbon monoxide, carbon dioxide, methane, black carbon, particulate matter, and Volatile Organic Compounds.
The month-long “integration test… lays the foundation for generating high resolution maps of air quality in cities,” Aclima says. The company is also partnered with the US Environmental Protection Agency, which says “mobile air measurements can complement existing stationary measurements for a more detailed picture of personal and community air quality.”
Nokia captures VR with spherical camera
Nokia may not be making phones anymore, but it’s not out of the hardware business: the company debuted a “virtual reality camera for professional content creators.”
The OZO is “the first commercially available virtual reality camera,” the company says. It was conceived at R&D facilities in Finland, and should ship by the end of the year.
It captures stereoscopic 3D video through eight synchronized global shutter sensors, and spatial audio through eight microphones. The system also enables real-time 3D viewing of the captured scene.
There’s more information here.
Face recognition in thermal images
Algorithms that identify people in photos are improving all the time, but a new technique takes recognition a few steps further: deep neural nets can now recognize your face in thermal images.
In other words, “face recognition technology that works in utter darkness,” Discovery reports.
The initial challenge is that “faces look different in the infrared and matching these images to their normal appearance is a significant unsolved challenge… the link between the way people look in infrared and visible light is highly nonlinear,” MIT Technology Review adds.
Researchers at the Karlsruhe Institute of Technology in Germany “worked out how to connect a mid- or far-infrared image of a face with its visible light counterpart for the first time.”
Well, they don’t do it themselves, of course — they’ve instead taught a neural network to identify the images.
“If perfected, face recognition technology could aid law enforcement with spotting known terrorists or other criminals walking through a mall or an airport,” Discovery adds.
Here is the full Discovery article.
Here is the full Technology Review article.
