'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 newsletter!

Or

Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.
Show Comments

Blog Posts

5 cornerstones of a strong digital culture

Creating a strong company culture may sound like a daunting task, but it’s actually pretty straightforward. In fact, company culture is created in exactly the same fashion as a religion or democracy. Behaviours created from the organisation’s inception are reinforced over time by leadership, attracting like-minded people and eventually reaching critical mass to become an accepted ‘truth’.

Anthony Stevens

Founder and CEO, Digital Asset Ventures

Should you rebuild your company’s tech stack in blockchain?

The question I get asked most regularly these days is: ‘Do I need to rebuild my company’s systems on Blockchain?’ And the answer, every time, is ‘No, you’re asking the wrong question’.

Michelle O'Keeffe

CEO, Engaging.io

Customer value proposition: Getting the brand promise to your customers right

Throughout my career, I have witnessed a litany of brand names that profess to have a unique customer value proposition (CVP). In reality, however, they’ve had little more than a ‘value proposition’: A simple list of benefits applied to a general audience.

Ric Navarro

Global director of marketing and communications, Norman, Disney & Young

We supply petroleum product such as JP54, D2, D6, JET A1, Serious buyer should contact now for SCO in order to work with seller workable ...

Vico Peißker

ACCC releases Consumer Data Right draft framework for comment

Read more

What is your opinion about chatbots serving your customers? My belief is bots are going to be the future of customer service and fulfilme...

Giridhar Prathap Reddy

NAB taps power of AI chatbots for business customer service

Read more

Thanks for writing about chatbots. Definitely bots have the exciting future when it comes to customer engagement, transactional and conve...

Giridhar Prathap Reddy

Deloitte TMT Predictions: AR and mobile crucial for marketing success in 2018

Read more

Hi there! Thanks for this interesting article! I love to read about new technology and software that makes our lives easier. I'm looking ...

Julia Summer

Wartsila overhauls Web platforms to create ‘seamless’ brand experience across all devices

Read more

RE: Sales and marketing SLAs, often the choke point isn't the teams but them getting the data into the tools they want to use with the da...

Ed Fry

Why sales and marketing alignment is more important than ever

Read more

Latest Podcast

More podcasts

Sign in