American scientists invent an ultra-fast camera with advanced computational vision

A team of researchers from the University of Washington and Princeton University has developed a new compact camera that relies on innovative techniques combining optics and computing, giving it advanced computational vision capabilities. The significance of this development lies in enabling devices to “perceive” visually by processing and analyzing images and videos in a way that mimics the human brain’s functionality.

The prototype of the new camera stands out for its ability to operate at the speed of light while consuming low energy, making it more efficient at recognizing and classifying objects.

The researchers replaced the traditional lens with a set of 50 meta-lenses, which are flat and lightweight optical components that use microscopic nanostructures to control light. These lenses were integrated to form a visual neural network, a type of artificial intelligence inspired by the brain’s function.

The researchers pointed out that this innovation offers two main advantages: the first is accelerating the process of image recognition and classification at a rate more than 200 times faster than traditional neural networks relying on computers, while maintaining the same level of accuracy. The second advantage is reducing energy consumption, as the technology relies on incoming light instead of electricity to perform the computational tasks.