moon_unit2 writes: From the story: You can, train a deep-learning algorithm to recognize a cat with a cat-fancier’s level of expertise, but you’ll need to feed it tens or even hundreds of thousands of images of felines, capturing a huge amount of variation in size, shape, texture, lighting, and orientation. It would be lot more efficient if, a bit like a person, an algorithm could develop an idea about what makes a cat a cat from fewer examples. A Boston-based startup called Gamalon has developed technology that lets computers do this in some situations, and it is releasing two products Tuesday based on the approach.
moon_unit2 writes: Apparently, playing computer games might provide a shortcut to greater intelligence. MIT Technology Review has a story about researchers using virtual game environments to train deep neural networks to recognize real-world objects. It's an important idea because deep learning usually requires huge quantities of annotated data, which isn't always available. So researchers from Xerox Europe, led by Adrien Gaidon, showed that training a deep learning system on a photo-realistic street scene could enable it to identify cars on real roads. “The nice thing about virtual worlds is you can create any kind of scenario,” Gaidon says. Perhaps video games could play a bigger role in the future of AI than anyone realized.
moon_unit2 writes: DJI's new Phantom 4 drone may be able to stop if there's an obstacles directly in front of it, but MIT Technology Review has a story about a much more sophisticated self-flying drone, from a startup called Skydio (basically using high-speed visual SLAM, which is no mean feat in such a tiny package). The company's prototype uses several video cameras to navigate around obstacles at high speeds through busy airspace. The technology could make consumer drones much harder to crash, and it could let drones do more complex surveillance tasks. Skydio, founded last year, has so far raised $25 million in funding in a round led by Andreessen Horowitz and Accel Partners.
moon_unit2 writes: You may have heard of "personal robots" such as Jibo, Buddy, and Pepper. One journalist recently met one of these home bots and found the reality less dazzling than the promotional videos. Whereas the Indiegogo clips of Buddy show the robot waking people up and helping with cooking, the current prototype can only perform a few canned tasks, and it struggles with natural language processing and vision. As the writer notes, the final version may be a lot more sophisticated, but it's hard to believe that real home helpers are just around the corner.
moon_unit2 writes: Amazon is organizing the event to spur the development of more nimble-fingered product-packing robots. Participating teams will earn points by locating products sitting somewhere on a stack of shelves, retrieving them safely, and then packing them into cardboard shipping boxes. Robots that accidentally crush a cookie or drop a toy will have points deducted. The contest is already driving new research on robot vision and manipulation, and it may offer a way to judge progress made in the past few years in machine intelligence and dexterity. Robots capable of advanced manipulation could eventually take on many simple jobs that are still done by hand.
An anonymous reader writes: Tired of bumping into paywalls while browsing big news sites? A new chrome extension, called Wrecking Ball, makes just about any paywall disappear by exploiting the way those companies give a free pass to anyone arriving from a search engine or through social media. So go enjoy NYT , WSJ, FT, Economist, and more, while they last.
moon_unit2 writes: MIT Technology Review has a story about BMW's new collaborative final-assembly-line robots. The move could prove a significant in the ongoing automation of work, as robots have previously been incapable of doing such jobs, and too dangerous to work in close proximity to humans. Robots like the ones at BMW’s South Carolina plant are also to cooperate with human workers, by handing them a wrench when they need it. So perhaps the next big shift in labor could be robot-human collaboration.
moon_unit2 writes: Technology Review has a piece on the reality behind all the hype surrounding self-driving, or driverless, cars. From the article: "Vehicle automation is being developed at a blistering pace, and it should make driving safer, more fuel-efficient, and less tiring. But despite such progress and the attention surrounding Google’s “self-driving” cars, full autonomy remains a distant destination. A truly autonomous car, one capable of dealing with any real-world situation, would require much smarter artificial intelligence than Google or anyone else has developed. The problem is that until the moment our cars can completely take over, we will need automotive technologies to strike a tricky balance: they will have to extend our abilities without doing too much for the driver."
moon_unit2 writes: Tech Review has a story about a garage in Ingolstadt, Germany, where the cars park themselves. The garage is an experiment set up by Audi to explore ways that autonomous technology might practically be introduced; most of the sensor technology is built into the garage and relayed to the cars rather than inside the cars themselves. It seems that carmakers see the technology progressing in a slightly different way to Google, with its fleet of self-driving Prius. From the piece: “It’s actually going to take a while before you get a really, fully autonomous car,” says Annie Lien, a senior engineer at the Electronics Research Lab, a shared facility for Audi, Volkswagen, and other Volkswagen Group brands in Belmont, California, near Silicon Valley. “People are surprised when I tell them that you’re not going to get a car that drives you from A to B, or door to door, in the next 10 years.”