Google has been granted a new patent that could potentially be pretty huge for the company. The patent is for automatic object recognition, which is usually quite difficult to pull off due to the need for human input. With Google’s patent, humans don’t need to do anything, as the system would make use of “feature vectors” to aid in the recognition of the objects in a video or still image.
Engadget reports that these feature vectors include things like color and movement to automatically recognize objects. These feature vectors are refined in the system’s learning process, where Google suggests that the system would use images from a “visual content repository” to become more familiar with the traits of the objects its meant to identify. It’s a very interesting patent, and the whole learning process could be helped along by content creators on YouTube who name and tag their videos accurately.
While it makes sense that Google would want to apply this patent to YouTube, it could also have some applications with Google Glass. Object recognition could be a huge feature of Google Glass should the company decide to implement something like this, so you can bet that Google is considering that avenue. At the moment, though, we don’t know how Google will utilize this freshly-granted patent, or if it even will to begin with.
Still, this could potentially lead to an object recognition system that’s completely automated, and that’s definitely something to get excited about. We’ll have to keep our eyes peeled for more information about this patent, but it might be a while before Google wants to talk about the system described in it. Stay tuned, folks.