News classification
Contact us
- Add: No. 9, North Fourth Ring Road, Haidian District, Beijing. It mainly includes face recognition, living detection, ID card recognition, bank card recognition, business card recognition, license plate recognition, OCR recognition, and intelligent recognition technology.
- Tel: 13146317170 廖经理
- Fax:
- Email: 398017534@qq.com
Artificial intelligence neural network to identify objects in the future or at the speed of light
Artificial intelligence neural network to identify objects in the future or at the speed of light
Artificial neural networks are created by 3D printers for medical, robotic and hacker safety. A team of electrical and computer engineers at UCLA created a physical artificial neural network, a device that mimics how the human brain works, capable of dissecting large amounts of data and identifying objects at the speed of light.
Today's many devices in everyday life use computerized cameras to identify objects, think about an ATM that can "read" the amount of handwritten money when depositing banknotes, or an Internet search engine that can quickly match photos to other similar images in their database. . But these systems rely on a device to stop imaging an object. First, "see" it with a camera or optical sensor, then dispose of the data it sees as data, and finally use a calculation program to figure out what it is.
The equipment developed by the University of California, Los Angeles is a step ahead. It is called the "diffractive depth neural network", which uses light reflected from the object itself to identify the object at the same time as the computer simply "sees" the object. The UCLA device does not require an advanced computing program to process the image of the object and decide what the object is after its optical sensor picks up. And there is no energy to run the device, since it only uses the diffraction of light.
New technologies based on the device can be used to speed up data-intensive tasks that touch and sequence objects. For example, a driverless car that uses this technology can respond immediately, even faster than using existing technology to respond to stop signs. Using a device based on the UCLA system, once the light from the marker hits it, the car will "read" the logo instead of having to "wait" the camera of the car to image the object and then use its computer to find out what the object is.
The technology based on this creation can also be used for microscopic imaging and medicine, for example, for sorting millions of cells for signs of disease. This optical artificial neural network device intuitively mimics how the brain processes information. It is able to extend the range to complete new camera designs and common optical components that work passively in medical technology, robotics, safety or any application where image and video data are essential.
Since its components can be created by 3D printers, artificial neural networks can be fabricated with larger and larger layers, resulting in devices with hundreds of millions of human-made neurons. Those larger devices can recognize more objects at the same time or perform more complex data profiling. And the components can be manufactured at low cost, and the equipment created by the UCLA team can be replicated below $50. Although the study uses light at terahertz frequencies, it can also create neural networks that use visible, infrared or other frequencies of light. The network can also be fabricated using lithography or other printing techniques.
Today's many devices in everyday life use computerized cameras to identify objects, think about an ATM that can "read" the amount of handwritten money when depositing banknotes, or an Internet search engine that can quickly match photos to other similar images in their database. . But these systems rely on a device to stop imaging an object. First, "see" it with a camera or optical sensor, then dispose of the data it sees as data, and finally use a calculation program to figure out what it is.
The equipment developed by the University of California, Los Angeles is a step ahead. It is called the "diffractive depth neural network", which uses light reflected from the object itself to identify the object at the same time as the computer simply "sees" the object. The UCLA device does not require an advanced computing program to process the image of the object and decide what the object is after its optical sensor picks up. And there is no energy to run the device, since it only uses the diffraction of light.
New technologies based on the device can be used to speed up data-intensive tasks that touch and sequence objects. For example, a driverless car that uses this technology can respond immediately, even faster than using existing technology to respond to stop signs. Using a device based on the UCLA system, once the light from the marker hits it, the car will "read" the logo instead of having to "wait" the camera of the car to image the object and then use its computer to find out what the object is.
The technology based on this creation can also be used for microscopic imaging and medicine, for example, for sorting millions of cells for signs of disease. This optical artificial neural network device intuitively mimics how the brain processes information. It is able to extend the range to complete new camera designs and common optical components that work passively in medical technology, robotics, safety or any application where image and video data are essential.
Since its components can be created by 3D printers, artificial neural networks can be fabricated with larger and larger layers, resulting in devices with hundreds of millions of human-made neurons. Those larger devices can recognize more objects at the same time or perform more complex data profiling. And the components can be manufactured at low cost, and the equipment created by the UCLA team can be replicated below $50. Although the study uses light at terahertz frequencies, it can also create neural networks that use visible, infrared or other frequencies of light. The network can also be fabricated using lithography or other printing techniques.