WASHINGTON, July 8 (Xinhua) — Engineers from University of Wisconsin-Madison (UW-Madison) demonstrated a “smart” mirror that can recognize images without any digital sensors or power sources.
The proof-of-concept research published on Monday in the journal Photonics Research revealed that one piece of glass may recognize your face without using any power at all in the future, and it used only analog material instead of digital codes.
Currently, the artificial intelligence tends to consume substantial computational resources, but the new invention offers a surprisingly “low-tech” approach.
The engineers embedded air bubbles of different sizes and shapes, and small pieces of light-absorbing materials like graphene at specific locations inside the glass that would bend light in specific ways, in order to differentiate among different images.
In one experiment, they devised a piece of glass that can identify handwritten numbers. Light emanating from an image of a number, such as “8,” enters at one end of the glass, and then focuses to one of nine specific spots on the other side, each corresponding to individual digits, according to the study.
Also, the glass is dynamic enough to detect, in real-time, when a handwritten “3” is altered to become an “8.”
“The wave dynamics of light propagation provide a new way to perform analog artificial neural computing,” said Yu Zongfu, professor of electrical and computer engineering at UW-Madison, who led the study.
The detection is completely passive and works at literally the speed of light. Also, one piece of image-recognition glass could be used hundreds of thousands of times, according to the study.
The researchers plan to determine if their approach works for more complex tasks, such as the biometric lock by recognizing one person’s face.