Even the most powerful computers are still no match for the human brain when it comes to pattern recognition, risk management, and other similarly complex tasks. Recent advances in optical neural networks, however, are closing that gap by simulating the way neurons respond in the human brain.
In a key step toward making large-scale optical neural networks practical, researchers have demonstrated a first-of-its-kind multilayer all-optical artificial neural network. Generally, this type of artificial intelligence can tackle complex problems that are impossible with traditional computational approaches, but current designs require extensive computational resources that are both time-consuming and energy intensive. For this reason, there is great interest developing practical optical artificial neural networks, which are faster and consume less power than those based on traditional computers.
In Optica, The Optical Society’s journal for high-impact research, researchers from The Hong Kong University of Science and Technology, Hong Kong detail their two-layer all-optical neural network and successfully apply it to a complex classification task.
“Our all-optical scheme could enable a neural network that performs optical parallel computation at the speed of light while consuming little energy,” said Junwei Liu, a member of the research team. “Large-scale, all-optical neural networks could be used for applications ranging from image recognition to scientific research.”
Building an all-optical network
In conventional hybrid optical neural networks, optical components are typically used for linear operations while nonlinear activation functions—the functions that simulate the way neurons in the human brain respond—are usually implemented electronically because nonlinear optics typically require high-power lasers that are difficult to implement in an optical neural network.
To overcome this challenge, the researchers used cold atoms with electromagnetically induced transparency to perform nonlinear functions. “This light-induced effect can be achieved with very weak laser power,” said Shengwang Du, a member of the research team. “Because this effect is based on nonlinear quantum interference, it might be possible to extend our system into a quantum neural network that could solve problems intractable by classical methods.”
To confirm the capability and feasibility of the new approach, the researchers constructed a two-layer fully-connected all optical neural network with 16 inputs and two outputs. The researchers used their all-optical network to classify the order and disorder phases of the Ising model, a statistical model of magnetism. The results showed that the all-optical neural network was as accurate as a well-trained computer-based neural network.
Optical neural networks at larger scales
The researchers plan to expand the all-optical approach to large-scale all-optical deep neural networks with complex architectures designed for specific practical applications such as image recognition. This will help demonstrate that the scheme works at larger scales.
“Although our work is a proof-of-principle demonstration, it shows that it may become possible in the future to develop optical versions of artificial intelligence,” said Du. “The next generation of artificial intelligence hardware will be intrinsically much faster and exhibit lower power consumption compared to today’s computer-based artificial intelligence,” added Liu.
The Latest on: Optical artificial neural networks
via Google News
The Latest on: Optical artificial neural networks
- Researchers build a silicon photonic neural network that overcomes limitations of fibre transmissionon January 16, 2022 at 2:30 am
Researchers at Princeton Lightwave Lab and NEC Laboratory America have built a real-time neural network on an integrated photonic chip, enabled by silicon photonics. The technology could be useful for ...
- The Carbon Skyscraperon January 12, 2022 at 4:00 pm
Yet there’s a built-in optical illusion that greatly ... To improve on that approach, my colleague Scott Kulp used neural networks, a form of artificial intelligence, to construct a continuous ...
- How simple liquids like water can perform complex calculationson January 11, 2022 at 5:14 am
After many decades of astonishing developments, advances in semiconductor-based computing are beginning to slow as transistors reach their physical limits in size and speed. However, the requirements ...
- Ions in the machine: How simple liquids like water can perform complex calculationson January 10, 2022 at 4:00 pm
To demonstrate the potential of chemical dynamics as a computing resource, researchers from Osaka University, the University of Tokyo, and Hokkaido University developed a method for building physical ...
- A silicon photonic-electronic neural network that could enhance submarine transmission systemson January 7, 2022 at 8:20 am
We are currently witnessing an explosion of network traffic. Numerous emerging services and applications, such as cloud services, video streaming platforms and the Internet of Things (IOT), are ...
- Tech notes from Day 1 of CES 2022on January 6, 2022 at 2:29 pm
CES 2022 may be down over 50% from 2020’s 170,000 in-person attendees, but the Consumer Technology Association (CTA) partnered with CLEAR and created an efficient badge pickup, ID and vaccination ...
- 3D holographic televisions are much closer than a galaxy far, far awayon January 5, 2022 at 3:02 pm
Pierre-Alexandre Blanche from the Wyant College of Optical Sciences at the University of Arizona ... especially in computation thanks to neural networks and artificial intelligence,” Blanche said.
- Ambarella (AMBA) Introduces CV3 AI Domain Controller SoCon January 5, 2022 at 9:30 am
Ambarella AMBA recently unveiled CV3 artificial intelligence (“AI ... operations per second of CVflow AI processing for neural network (“NN”) computation. The system includes a general ...
via Bing News