From M.I.T Technology Review: “Google Researchers Have a New Alternative to Traditional Neural Networks”

MIT Technology Review
M.I.T Technology Review

November 1st, 2017
Jamie Condliffe

1
Image credit: Jingyi Wang

Say hello to the capsule network.

AI has enjoyed huge growth in the past few years, and much of that success is owed to deep neural networks, which provide the smarts behind some of AI’s most impressive tricks like image recognition. But there is growing concern that some of the fundamental tenets that have made those systems so successful may not be able to overcome the major problems facing AI—perhaps the biggest of which is a need for huge quantities of data from which to learn.

Seemingly Google’s Geoff Hinton is among those who are concerned. Because Wired reports that he has now unveiled a new take on the traditional neural networks that he calls capsule networks. In a pair of new papers—one published on the arXIv, the other on OpenReview—Hinton and a handful of colleagues explain how they work.

Their approach uses small groups of neurons, collectively known as capsules, which are organized into layers to identify things in video or images. When several capsules in one layer agree on having detected something, they activate a capsule at a higher level—and so on, until the network is able to make a judgement about what it sees. Each of those capsules is designed to detect a specific feature in an image in such a way that it can recognize them in different scenarios, like from varying angles.

Hinton claims that the approach, which has been in the making for decades, should enable his networks to require less data than regular neural nets in order to recognize objects in new situations. In the papers published so far, capsule networks have been shown to keep up with regular neural networks when it comes to identifying handwritten characters, and make fewer errors when trying to recognize previously observed toys from different angles. In other words, he’s published the results because he’s got his capsules to work as well as, or slightly better than, regular ones (albeit more slowly, for now).

Now, then, comes the interesting part. Will these systems provide a compelling alternative to traditional neural networks, or will they stall? Right now it’s impossible to tell, but we can expect the machine learning community to implement the work, and fast, in order to find out. Either way, those concerned about the limitations of current AI systems can be heartened by the fact that researchers are pushing the boundaries to build new deep learning alternatives.

See the full article here .

Please help promote STEM in your local schools.

STEM Icon

Stem Education Coalition

The mission of MIT Technology Review is to equip its audiences with the intelligence to understand a world shaped by technology.