Microsoft's imaging tech is (sometimes) better than you at spotting objects

Many computer vision projects struggle to mimic what people can achieve, but Microsoft Research thinks that its technology might have already trumped humanity... to a degree, that is. The company has published results showing that its neural network technology made fewer mistakes recognizing objects than humans in an ImageNet challenge, slipping up on 4.94 percent of pictures versus 5.1 percent for humans. One of the keys was a "parametric rectified linear unit" function (try saying that three times fast) that improves accuracy without any real hit to processing performance.

You aren't about to get many keen-sighted artificial intelligences just yet. Microsoft is quick to note that its vision system (like others) excels in tests like these, where there are subtle distinctions that flesh-and-bone observers can't always see. Computers are more likely to goof up with simpler recognition tasks, like identifying barnyard animals. Still, it's noteworthy that software emerged victorious in the first place. Besides leading to smarter photo software that's better at organizing your vacation shots, the improved detection should help autonomous robots and other devices that need to make real-life snap decisions based purely on what they see.