While we are getting used to unlock our phones using facial recognition, efforts are being made to identify individuals detected in the dark. The proposed system does so by using infrared scanning methodology to generate thermal heat map of person to be detected and then comparing it against a standard photograph of him. Such thermal sensors are expected to be successful in varied environments even when targets:
- are blocked by automotive windshield glasses.
- are obscured due to weather conditions (like rain, fog, snow).
- are spotted far away.
- have extremely bright background light.
What makes this invention so challenging is that there’s little correlation between how you appear in the daylight versus how you appear on an infrared camera. To complicate it further, if you’ve been running or have a fever, your face’s map is going to look different as well. All of this makes it a very tricky problem to solve. To overcome this, the system had to be fed with many different sample images of people in different lighting, pulling different facial expressions and at different times of the day. With enough pictures to refer to, the model could accurately spot a face 80 percent of the time, and in just 35 milliseconds!
The mentioned project is being funded by US Military’s Department of Defense, which expects the technology to be incorporated in a device that is small enough to be carried by an individual. It should be able to operate from a distance of 10 to 500 meters and match individuals against a watch list. Once developed and refined, it could help spot criminals in the dark, whether driving past a speed camera or rushing through a building at night.
For sure, the topic is bound to raise questions regarding privacy but like many emerging technologies, a balance needs to be struck between keeping us safe versus having our every move logged and recorded.
This content was originally published for my TechTuesday’s initiative on LinkedIn.
