The US military is developing a technology that can shoot through the wall and can distinguish from outside who is in the room

The US Army is developing a technology to discriminate people by machine learning images of infrared thermal cameras. Once this technology is completed, it will be possible to even photograph indoors from outside to identify the person inside.

Thermal to Visible Synthesis of Face Images using Multiple Regions
(PDF file)https://arxiv.org/pdf/1803.07599.pdf

The US Army is developing AI that can recognize faces in the dark and through walls
https://thenextweb.com/artificial-intelligence/2018/04/18/the-us-army-is-developing-ai-that-can-recognize-faces-in-the-dark-and-through-walls/

A research team at Army Research Laboratory, Mr. Benjamin Rigan, is studying how to apply face recognition technology to images taken with a thermal camera. In the research, a technique to identify features of parts of facial images photographed with thermal cameras using machine learning with known face photograph lists has been developed.


Research seems to be still in the early stages of technology development, but we extracted the features of the thermal image that photographed the person's face and succeeded in identifying a person from the face picture list taken with a normal camera about. In the future, it is aimed to realize real-time face recognition at the work site by making it possible to check thermal images taken through obstacles such as walls that hinder visibility against the facial photograph list .


When this technology is completed, it is of course useful as a military technology, but since it is possible to judge who is in the room from outside regardless of day and night, if it is used for surveillance cameras and general investigation methods privacy It is unlikely that it will be a major threat to.