The “invisibility cloak” was invented in China
Students in China have invented an “invisibility cloak” designed to hide the wearer from security cameras. InvisDefense uses an unusual pattern and erratic heat signals to confuse AI cameras
Developed by a team from Wuhan University in China, InvisDefense is a low-cost jacket that can be used day and night to avoid detection by AI-tracked security camera systems.
The jacket uses a pattern to effectively blind cameras during the day, while at night it emits unusual heat signals to confuse infrared cameras.
The jacket, which looks ordinary to the human eye, is ineffective at hiding users from human-watched security cameras. The design of the jacket means that some sensing systems used in self-driving vehicles can also confuse human drivers, but not human drivers.
CANNOT BE DETECTED IF HUMAN
“Today, many surveillance devices can detect the human body,” Professor Vang Zheng of the Wuhan University School of Computer Science, who pioneered the invention, told the South China Morning Post.
Cameras on the road have pedestrian detection functions, and smart cars can identify pedestrians, roads and obstacles. InvisDefense allows the camera to capture you, but the camera cannot tell if you are human or not.
An article describing the technology used in the jacket, which won the first prize in a competition sponsored by Huawei Technologies, is planned to be presented at the AAAI 2023 AI conference to be held in the USA in February.
CHEAP THAN 1400 LIRA
Wuhan University doctoral student Vei Hui, who created the algorithm that creates the pattern of the jacket, calculated that it costs less than 500 yuan to make InvisDefense.
“The hardest part is the balance of the camouflage pattern. Traditionally, researchers used bright images to block machine vision, and it worked,” said Vei Hui.
However, this attracts the human eye, making the user even more remarkable. We use algorithms to design the least noticeable patterns that can disable computer vision.
Professor Vang said the jacket could be used for military camouflage to avoid detection by AI-controlled units or drones on the battlefield.
Researchers could also potentially use the model-building algorithm to better understand and fix gaps in artificial intelligence and computer recognition technology.