'Ex Machina', here we come: A new algorithm helps computers learn the way we do

In a 'visual Turing test,' human judges couldn't tell them apart

Machine learning is all about getting computers to "understand" new concepts, but it's still a pretty inefficient process, often requiring hundreds of examples for training. That may soon change, however, thanks to new research published on Friday.

Aiming to shorten the learning process and make it more like the way humans acquire and apply new knowledge based on just a few examples, a team of researchers has developed what they call a Bayesian Program Learning framework and then used it to teach computers to identify and reproduce handwritten characters based on just a single example.

Whereas standard pattern-recognition algorithms represent concepts as configurations of pixels or collections of features, the BPL approach learns by “explaining” the data provided to the algorithm -- in this case, the sample character. Concepts are represented as probabilistic computer programs and the algorithm essentially programs itself by constructing code to produce the letter it sees. It can also capture variations in the way different people draw a given letter.

The model also “learns to learn” by using knowledge from previous concepts to speed learning on new ones, so it can use knowledge of the Latin alphabet to learn letters in the Greek alphabet more quickly, for example.

Most compelling of all is that the algorithm allowed computers to pass a sort of "visual Turing test." Specifically, the researchers asked both humans and computers to reproduce a series of handwritten characters after being shown just a single example of each; in some cases, subjects were asked to create entirely new characters in the style of those originally shown. Bottom line: human judges couldn't tell the results apart.

The researchers have applied their model to more than 1,600 types of handwritten characters in 50 writing systems, including Sanskrit, Tibetan, Gujarati and Glagolitic. They even tried it on invented characters such as those from the television series "Futurama."

A paper describing the research was published Friday in the journal Science. Its authors were Brenden Lake, a Moore-Sloan Data Science Fellow at New York University; Ruslan Salakhutdinov, an assistant professor of Computer Science at the University of Toronto; and Joshua Tenenbaum, a professor at MIT in the Department of Brain and Cognitive Sciences and the Center for Brains, Minds and Machines.

“It has been very difficult to build machines that require as little data as humans when learning a new concept,” said Salakhutdinov. “Replicating these abilities is an exciting area of research connecting machine learning, statistics, computer vision, and cognitive science.”

Join the CMO newsletter!

Error: Please check your email address.
Show Comments

Supporting Association

Blog Posts

Un-complicating multi-channel marketing: 5 actionable steps

There’s so much choice available that customers can pick and choose who they buy from and where, when, and how it happens. They want to discover, research, evaluate, and purchase on their preferred channel. Give them that option, and they’re more likely to choose you. That’s the whole point behind the multi-channel approach.

Aaron Agius

Co-founder and managing director, Louder Online

People in vegan houses shouldn't throw bacon

Picture this. You’re at a Gourmerican burger joint chomping a cheeseburger, when an outspoken vegan friend starts preaching that you’re killing the planet. Last week, that same vegan downed a pricey glass of pinot before their flight to a far-flung destination, armed with their strongest mossie repellant and first aid kit. Anything amiss?

Abbie Love

Strategist, Ikon Communications

The role of the CMO is evolving: Are you keeping up?

My (amazing) vacation in the Galapagos Islands earlier in the year got me thinking about Charles Darwin and his theory of evolution. What does this have to do with the role of today’s CMO, you ask? Plenty.

Sheryl Pattek

Vice-president, executive partner

Very good article, Social media analytics helps in problem identification. They can serve as an early warning system for negative custome...


Four ways to use social media to boost customer loyalty

Read more

It’s excellent aiming to resurrect the complete within the hearts and minds of connected customers, moreover because the terribly relevan...


CMO Interview: How Kodak’s global CMO is bringing the brand back from the brink

Read more

Great to see ActiveCampaign's growth funded with some serious money.As a platform, it's up there with the usual suspects in terms of feat...

Lawrence Ladomery

CMO's top 10 martech stories for the week - 13 October

Read more


Kerry Edwards

Open Colleges taps into social for better student interaction

Read more

Or just go to sites like www.shopsthatshiptoaustralia.c... and others and be sure that the stores will send to where you live :-)


Why online shopping is like dating – RedBalloon CEO

Read more

Latest Podcast

More podcasts

Sign in