“Hey Siri, how’s my hair?”
Your smartphone may soon be able to give you an honest answer, thanks to a new machine learning algorithm designed by U of T Engineering researchers Parham Aarabi (ECE) and Wenzhi Guo (ECE MASc 1T5).
The team designed an algorithm that learns directly from human instructions, rather than an existing set of examples, and outperformed conventional methods of training neural networks by 160 per cent. But more surprisingly, their algorithm also outperformed its own training by nine per cent — it learned to recognize hair in pictures with greater reliability than that enabled by the training, marking a significant leap forward for artificial intelligence.
Aarabi and Guo trained their algorithm to identify people’s hair in photographs — a much more challenging task for computers than it is for humans.
“Our algorithm learned to correctly classify difficult, borderline cases — distinguishing the texture of hair versus the texture of the background,” says Aarabi. “What we saw was like a teacher instructing a child, and the child learning beyond what the teacher taught her initially.”
Humans “teach” neural networks — computer networks that learn dynamically — by providing a set of labeled data and asking the neural network to make decisions based on the samples it’s seen. For example, you could train a neural network to identify sky in a photograph by showing it hundreds of pictures with the sky labeled.
This algorithm is different: it learns directly from human trainers. With this model, called heuristic training, humans provide direct instructions that are used to pre-classify training samples rather than a set of fixed examples. Trainers program the algorithm with guidelines such as “Sky is likely to be varying shades of blue,” and “Pixels near the top of the image are more likely to be sky than pixels at the bottom.”
Their work is published in the journal IEEE Transactions on Neural Networks and Learning Systems.
This heuristic training approach holds considerable promise for addressing one of the biggest challenges for neural networks: making correct classifications of previously unknown or unlabeled data. This is crucial for applying machine learning to new situations, such as correctly identifying cancerous tissues for medical diagnostics, or classifying all the objects surrounding and approaching a self-driving car.
“Applying heuristic training to hair segmentation is just a start,” says Guo. “We’re keen to apply our method to other fields and a range of applications, from medicine to transportation.”
The Latest on: AI algorithm heuristic training
via Google News
The Latest on: AI algorithm heuristic training
- Facebook details self-supervised AI that can segment images and videoson April 30, 2021 at 2:00 pm
Facebook's DINO, an AI system, can train a model to segment an image or video without training on a labeled dataset.
- Artificial Intelligence Algorithm Helps Unravel the Physics Underlying Quantum Systemson April 30, 2021 at 2:15 am
Protocol to reverse engineer Hamiltonian models advances automation of quantum devices. Scientists from the University of Bristol's Quantum Engineering Technology Labs (QETLabs) have developed an algo ...
- Inspur Rises On The Wave of AI Servers In The Datacenteron April 29, 2021 at 8:10 am
Machine learning techniques based on massive amounts of data and GPU accelerators to chew through it are now almost a decade old. They were originally ...
- The Artificial Intelligence Explosion in Industrial Automationon April 28, 2021 at 11:36 am
From supply chain and workforce training to automated quality inspections, artificial intelligence capabilities are being added to automation technologies at a rapid pace and are changing not only ...
- DARPA Looking to Infuse Aerial Systems with AIon April 27, 2021 at 7:02 pm
The Defense Advanced Research Projects Agency wants to help the military incorporate more artificial intelligence capabilities into air combat systems in the coming years, officials recently said.
- NLR is Testing Flight Schedule Driven AI for Commercial Aircraft Maintenance Planningon April 26, 2021 at 3:45 pm
Amsterdam-based applied science institute Netherlands Aerospace Laboratory (NLR) is in discussion with several airlines about the ongoing development of its artificial intelligence algorithms ... (DP) ...
- Rice Univ. Researchers Claim 15x AI Model Training Speed-up Using CPUson April 23, 2021 at 3:05 pm
Reports are circulating in AI circles that researchers from Rice University claim a breakthrough in AI model training acceleration – without using accelerators. Running AI software on commodity x86 ...
- Hitting the Books: How biased AI can hurt users or boost a business's bottom lineon April 10, 2021 at 9:35 am
Or rather, perhaps we humans must first detangle ourselves from these very same biases before expecting them eliminated from our algorithms. In A Citizen's Guide to Artificial Intelligence ...
- Why AI can’t solve unknown problemson April 2, 2021 at 2:40 am
Meanwhile, the AI toolbox continues to grow with algorithms that can perform specific ... these tasks with much effort and often a lot of training, but we should be suspicious if we think that ...
via Bing News