Researchers at UCLA used a 3D printer to create an artificial neural network capable of analyzing large volumes of data and identifying objects at the speed of light. The system is called deep diffractive neural network (D2NN). It uses the light scattered by an object to identify the object.
UCLA researchers based the system on a project based on deep learning using passive diffractive layers working together. The researchers first created a computer-simulated design and then used a 3D printer to create thin, eight-inch square polymer inserts. Each of these cookies has uneven surfaces to help diffract light from an object.
3D printed wafers are penetrated using terahertz frequencies. Each layer is made up of tens of thousands of pixels through which light can travel. The design has each type of object assigned to a pixel, with the light coming from the object diffracted towards the pixel that has been assigned to its type. The technique allows D2NN to identify an object in the same amount of time as a computer would take to see the object.
The network was trained to learn the diffracted light that each object produced as the light from that object passes through the device using a branch of AI called deep learning. Deep learning teaches machines through repetition and over time, as patterns emerge. During the experiments, the device was able to accurately identify handwritten numbers and items of clothing.
The device has also been trained to act as an image lens, similar to the operation of a typical camera lens. As the device is created on a 3D printer, the D2NN can be made with larger and additional layers resulting in a device with hundreds of millions of artificial neurons. Larger devices can find many more objects simultaneously with the potential to perform more complex data analysis. Another critical aspect of D2NN is the cost, with the researcher saying that the device could be reproduced for less than $ 50.