As IFA was winding down on Friday, Google released a blog post detailing its latest object recognition technology, which Google research team ‘GoogLeNet’ used to place first in the classification and detection categories of the world’s largest academic competition for computer vision: the ImageNet large-scale visual recognition challenge.
The improved technology is said to be able to not only quickly and effectively recognize objects in images, it can also locate and label multiple object sizes in one image and can determine an object within or on top of an object. Google points out that these tech advances could allow better image understanding and could make its way to photo/image search, Youtube, self-driving cars or “any place where it is useful to understand what is in an image as well as where things are”. While Google doesn’t mention it specifically, we could also see how such tech could be very useful in robotics and similar fields.
Of course, Google isn’t keeping its latest breakthroughs to itself. Google’s reach team participated at the ILSVRC on an open submission, which means the exact details of the approach were shared with the computer vision community. For more in-depth details on Google’s approach, you’ll want to check out their official blog post.