Technique should benefit smart phones and self-driving cars
Whetstone, a software tool that sharpens the output of artificial neurons, has enabled neural computer networks to process information up to a hundred times more efficiently than the current industry standard, say the Sandia National Laboratories researchers who developed it.
The aptly named software, which greatly reduces the amount of circuitry needed to perform autonomous tasks, is expected to increase the penetration of artificial intelligence into markets for mobile phones, self-driving cars and automated interpretation of images.
“Instead of sending out endless energy dribbles of information,” Sandia neuroscientist Brad Aimone said, “artificial neurons trained by Whetstone release energy in spikes, much like human neurons do.”
The largest artificial intelligence companies have produced spiking tools for their own products, but none are as fast or efficient as Whetstone, says Sandia mathematician William Severa. “Large companies are aware of this process and have built similar systems, but often theirs work only for their own designs. Whetstone will work on many neural platforms.”
The open-source code was recently featured in a technical article in Nature Machine Intelligence and has been proposed by Sandia for a patent.
How to sharpen neurons
Artificial neurons are basically capacitors that absorb and sum electrical charges they then release in tiny bursts of electricity. Computer chips, termed “neuromorphic systems,” assemble neural networks into large groupings that mimic the human brain by sending electrical stimuli to neurons firing in no predictable order. This contrasts with a more lock-step procedure used by desktop computers with their pre-set electronic processes.
Because of their haphazard firing, neuromorphic systems often are slower than conventional computers but also require far less energy to operate. They also require a different approach to programming because otherwise their artificial neurons fire too often or not often enough, which has been a problem in bringing them online commercially.
Whetstone, which functions as a supplemental computer code tacked on to more conventional software training programs, trains and sharpens artificial neurons by leveraging those that spike only when a sufficient amount of energy — read, information —has been collected. The training has proved effective in improving standard neural networks and is in process of being evaluated for the emerging technology of neuromorphic systems.
Catherine Schuman, a neural network researcher at Oak Ridge National Laboratories, said, “Whetstone is an important tool for the neuromorphic community. It provides a standardized way to train traditional neural networks that are amenable for deployment on neuromorphic systems, which had previously been done in an ad hoc manner.”
The strict teacher
The Whetstone process, Aimone said, can be visualized as controlling a class of talkative elementary school students who are tasked with identifying an object on their teacher’s desk. Prior to Whetstone, the students sent a continuous stream of sensor input to their formerly overwhelmed teacher, who had to listen to all of it — every bump and giggle, so to speak — before passing a decision into the neural system. This huge amount of information often requires cloud-based computation to process, or the addition of more local computing equipment combined with a sharp increase in electrical power. Both options increase the time and cost of commercial artificial intelligence products, lessen their security and privacy and make their acceptance less likely.
Under Whetstone, their newly strict teacher only pays attention to a simple “yes” or “no” measurement of each student — when they raise their hands with a solution, rather than to everything they are saying. Suppose, for example, the intent is to identify whether a piece of green fruit on the teacher’s desk is an apple. Each student is a sensor that may respond to a different quality of what may be an apple: Does it have the correct quality of smell, taste, texture and so on? And while the student who looks for red may vote “no” the other student who looks for green would vote “yes.” When the number of answers, either yay or nay, is electrically high enough to trigger the neuron’s capacity to fire, that simple result, instead of endless waffling, enters the overall neural system.
While Whetstone simplifications could potentially increase errors, the overwhelming number of participating neurons — often over a million — provide information that statistically make up for the inaccuracies introduced by the data simplification, Severa said, responsible for the mathematics of the program.
“Combining overly detailed internal information with the huge number of neurons reporting in is a kind of double booking,” he says. “It’s unnecessary. Our results tell us the classical way — calculating everything without simplifying — is wasteful. That is why we can save energy and do it well.”
Patched programs work best
The software program works best when patched in to programs meant to train new artificial-intelligence equipment, so Whetstone doesn’t have to overcome learned patterns with already established energy minimums.
The Latest on: Artificially intelligent devices
via Google News
The Latest on: Artificially intelligent devices
- Synaptics integration could enable voice biometrics in low-power edge deviceson October 11, 2021 at 8:58 am
Synaptics and Edge Impulse partner to facilitate the development of embedded machine learning models in power-sensitive intelligent IoT applications.
- Artificial Intelligence Chip Market size worth $ 109.83 Billion, Globally, by 2028 at 40.17% CAGR: Verified Market Research®on October 11, 2021 at 8:15 am
Verified Market Research recently published a report, "Artificial Intelligence Chip Market" By End-user (Healthcare, Manufacturing, ...
- Robotic Assistance Devices to Reveal RAD 3.0 via YouTube Live Stream October 13on October 11, 2021 at 6:00 am
Artificial Intelligence Technology Solutions, Inc., (OTCPK:AITX), today announced that its wholly-owned subsidiary Robotic Assistance Devices Inc. (RA ...
- The Ethics of Artificial Intelligence (AI)on October 11, 2021 at 1:14 am
Ethical AI works to ensure that AI initiatives do not cause harm in any way to people. Explore the AI ethical issues & standards: ...
- Patents and Artificial Intelligence: An ‘Obvious’ Slippery Slopeon October 8, 2021 at 5:03 am
Stephen Thaler and Ryan Abbott plan to bring a light beacon, a beverage container, and a machine called Dabus into court, along with a simple question: Does an inventor need to be human?
- FDA Review Can Limit Bias Risks in Medical Devices Using Artificial Intelligenceon October 6, 2021 at 5:00 pm
Artificial intelligence (AI)—broadly defined as the science of developing machines that can simulate human thinking—offers great potential to improve health care and reduce costs.
- Google tries to make traffic lights more efficient with artificial intelligenceon October 6, 2021 at 10:00 am
Alphabet Inc's Google cut fuel use and traffic delays by 10% to 20% at four locations in Israel by using artificial intelligence to optimize signal lights and it next plans ...
- Artificial intelligence aids colonoscopies at Virginia VA medical centeron October 5, 2021 at 10:50 am
Virginia VA became the first hospital in the U.S. to use the GI Genius intelligent endoscopy module for colorectal screenings for Veterans.
- How Artificial Intelligence can change Educationon September 30, 2021 at 12:17 pm
Conversations about Artificial Intelligence usually involve robots, space exploration machines, and smart home devices. The impact of this technology on education is much less discussed, although AI ...
- Artificial Intelligence in Manufacturing Industry is Expected to Reach US$ 11.5 Bn by 2027on September 30, 2021 at 8:00 am
The Artificial Intelligence in manufacturing market is forecasted to grow at a high rate owing to the accelerating innovations in industrial IoT and automation.PLEASANTON CA, Sept. 30, 2021 (GLOBE ...
via Bing News