Meet the algorithm that can learn “everything about anything”

Summary:

Researchers from Allen Institute for AI have built a computer system capable of teaching itself many facets of broad concepts by scouring and analyzing search engines using natural language processing and computer vision techniques.

The most recent advances in artificial intelligence research are pretty staggering, thanks in part to the abundance of data available on the web. We’ve covered how deep learning is helping create self-teaching and highly accurate systems for tasks such as sentiment analysis and facial recognition, but there are also models that can solve geometry and algebra problems, predict whether a stack of dishes is likely to fall over and (from the team behind Google’s word2vec) understand entire paragraphs of text.

(Hat tip to frequent commenter Oneasum for pointing out all these projects.)

One of the more interesting projects is a system called LEVAN, which is short for Learn EVerything about ANything and was created by a group of researchers out of the Allen Institute for Artificial Intelligence and the University of Washington. One of them, Carlos Guestrin, is also co-founder and CEO of a data science startup called GraphLab. What’s really interesting about LEVAN is that it’s neither human-supervised nor unsupervised (like many deep learning systems), but what its creators call “webly supervised.”

What that means, essentially, is that LEVAN uses the web to learn everything it needs to know. It scours Google Books Ngrams to learn common phrases associated with a particular concept, then searches for those phrases in web image repositories such as Google Images, Bing and Flickr. For example, LEVAN now knows that “heavyweight boxing,” “boxing ring” and “ali boxing” are all part of the larger concept of “boxing,” and it knows what each one looks like.

More impressive still is that because LEVAN uses text and image references to teach itself concepts, it’s also able to learn when words or phrases mean the same thing. So while it might learn, for example, that “Mohandas Gandhi” and “Mahatma Gandhi” are both sub-concepts of “Gandhi,” it will also learn after analyzing enough images that they’re the same person.

So far, LEVAN has modeled 150 different concepts and more than 50,000 sub-concepts, and has annotated more than 10 million images with information about what’s in them and what’s happening in them. The project website lets you examine its findings for each concept and download the models.

According to a recent presentation by one of its creators, LEVAN was designed to run nicely on the Amazon Web Services cloud — yet another sign of how fast the AI space is moving. Computer science skills and math knowledge are one impediment to broadly accessible AI, but those can be addressed by SDKs, APIs, and other methods of abstracting complexity. However, training AI models can require a lot of computing power, something that is easily available to the likes of Facebook and Google but that for everyday users might need to be offloaded to the cloud.

ORIGINAL: GigaOM
By Derrick Harris
May. 23, 2014 - 10:16 AM PDT

Neuroscientists reach major milestone in whole-brain circuit mapping project

ORIGINAL: Cold Springs Harbor Laboratory

May 31, 2012

Projections from a motor cortex AAV injection (credit: CSHL)

Neuroscientists at at Cold Spring Harbor Laboratory(CSHL) reached an important milestone today, publicly releasing the first installment of data from the 500 terabytes so far collected in their pathbreaking project to construct the first whole-brain wiring diagram of a vertebrate brain, that of the mouse.

Cold Spring Harbor, NY – Neuroscientists at Cold Spring Harbor Laboratory (CSHL) reached an important milestone today, publicly releasing the first installment of data from the 500 terabytes so far collected in their pathbreaking project to construct the first whole-brain wiring diagram of a vertebrate brain, that of the mouse. 

Add caption

The data consist of gigapixel images (each close to 1 billion pixels) of whole-brain sections that can be zoomed to show individual neurons and their processes, providing a “virtual microscope.” The images are integrated with other data sources from the web, and are being made fully accessible to neuroscientists as well as interested members of the general public (http://mouse.brainarchitecture.org). The data are being released pre-publication in the spirit of open science initiatives that have become familiar in digital astronomy (e.g., Sloan Digital Sky Survey) but are not yet as widespread in neurobiology. 

Each sampled brain is represented in about 500 images, each image showing an optical section through a 20 micron-thick slice of brain tissue. A multi-resolution viewer permits users to journey through each brain from “front” to “back,” and thus enables them to follow the pathways taken through three-dimensional brain space by tracer-labeled neuronal pathways. The tracers were picked to follow neuronal inputs and outputs of given brain regions.

“We’re executing a grid-based “shotgun” strategy for neuronal tract tracing that we first proposed a few years ago, and which I am pleased to note has gained acceptance elsewhere within the neuroscience community,” says Partha P. Mitra, Ph.D., the Crick-Clay Professor of Biomathematics at CSHL and director of the Mouse Brain Architecture (MBA) Project. After the initial June 1 release, project data will be made public continuously on a monthly basis, Mitra says.

Project addresses a large gap in knowledge

“Our project seeks to address a remarkable gap in our knowledge of the brain,” Mitra explains. “Our knowledge of how the brain is wired remains piecemeal and partial after a century of intense activity. Francis Crick and Ted Jones emphasized this in an article published in Nature nearly 20 years ago. Yet to understand how the brain works (or fails to work in neurological or neuropsychiatric disease), it is critical that we understand this wiring diagram more fully. Further, there remain fundamental questions about brain evolution that cannot be addressed without obtaining such wiring diagrams for the brains of different species.”