Herself’s Artificial Intelligence

Infants still have a chubby leg up on the best supercomputers when it comes to picking up languages, but now researchers have created a program that can teach itself to distinguish vowel sounds like those in “train” and “bed.” The software, the first to figure out vowel categories from human sounds, lends credence to the notion that language is more learned than innate.

. . .
Instead of telling his neural network ahead of time how many categories to expect, the researchers had their network constantly guess how many categories there ought to be while analyzing thousands of sound clips. The program quickly began clustering the sounds into only a few vowel categories. It could lump the vowel sounds from the average mother into four categories more than 80% of the time, McClelland’s team reports online this week in the Proceedings of the National Academy of Sciences. And that, says McClelland, bolsters the idea that much can be learned with little assumed. . . .

More information:
A baby step for computer learning

The Lockheed Martin Multifunction Utility/Logistics and Equipment (MULE) robotic vehicle’s Engineering Evaluation Unit (EEU) recently reached a major milestone in demonstrating autonomous navigation over complex obstacles, such as steps and gaps. The EEU autonomously climbed a 30-inch step and bridged a 70-inch gap without operator intervention, using only parametric descriptions of the obstacles and the vehicle’s self-awareness.

This capability exceeds the performance of other high-mobility vehicles, such as the HMMWV. Although a smaller vehicle, the MULE is able to address complex obstacles, such the ones used for the demonstration at a testing facility, by employing its specialized articulating suspension.

More information:
Lockheed Martin Reaches Major Milestone For the Mule Robotic Vehicle Engineering Unit
Multifunction Utility/ Logistics and Equipment Vehicle

The game of checkers has roughly 500 billion billion possible positions (5 x 1020). The task of solving the game, determining the final result in a game with no mistakes made by either player, is daunting. Since 1989, almost continuously, dozens of computers have been working on solving checkers, applying state-of-the-art artificial intelligence techniques to the proving process. This paper announces that checkers is now solved: perfect play by both sides leads to a draw. This is the most challenging popular game to be solved to date, roughly one million times more complex than Connect Four. Artificial intelligence technology has been used to generate strong heuristic-based game-playing programs, such as DEEP BLUE for chess. Solving a game takes this to the next level, by replacing the heuristics with perfection.

Test your skills at http://www.cs.ualberta.ca/~chinook/New. The user name is ‘checkers’, and the password is ‘solved2007′.

More information:
Chinook – World Man-Machine Checkers Champion
Paper at Science for a fee

Beneath the white paperboard petals of a robotic flower–which can open and close in response to changes in light, or catch a thrown ball detected by infrared sensors–lies a new standardized robotics platform called Qwerk. Developed at Carnegie Mellon University (CMU), Qwerk is designed so that almost anyone can use it to build his or her own custom Internet-enabled robot. It’s a platform that CMU computer scientist Illah Nourbakhsh hopes will launch an open-source robotics movement and “democratize robot design for people intimidated by current techniques and parts.”

In contrast to current kits–most of which require a prefabricated set of parts–Qwerk is, according to the CMU robotics team, the first easy-to-use, low-cost robotics controller to house, in one place, power regulators, motor controllers, and hardware and rewritable software for a Wi-Fi Internet connection and simple programming. In the flower robot, the platform sits inside the blue wooden flowerpot. The CMU team has also developed some robot recipes for easy-to-build machines–like the paperboard flower–that can be assembled in a few hours with off-the-shelf parts. Together, the recipes and platform make up the Telepresence Robotic Kit (TeRK). . .

More information:
Democratizing Robot Design | Technology Review
TeRK Telepresence Robot Kit