New Type Of Artificial Eye Recognises Images in Nanoseconds
Researchers from the Vienna University of Technology have presented a new type of artificial eye that combines light-sensing electronics with a neural network on a single chip with a capacity to preprocess data in nanoseconds, before sending it to the brain.
While usually computer vision required high computational power, new device can capture and process data at the same time, which “makes image recognition much faster using much less power”, shared MIT Technology Review.
In the article “Ultrafast machine vision with 2D material neural network image sensors” published in Nature on March 4, researchers have shared that their artificial eye can preprocess information before its further processing to the brain:
“We demonstrate that an image sensor can itself constitute an ANN [artificial neural network] that can simultaneously sense and process optical images without latency”
With the limited vision of the current artificial eye, researchers plan to scale the neural network, so the device will be able to capture images of larger sizes.
Image Credit: MIT Technology Review