Google has placed first in a competition to find automated technologies to recognise elements or locations in images. The GoogleNet team was top in the classification and detection tasks, which basically means its software algorithms accurately assign labels to a picture.

The detection challenge requires the ability for a computer to study a complex scene and correctly identify the elements in it. Examples given by Google include a bag of fruit, where each item is correctly assigned a tag, and a living room scene where even the pet cat is identified.

Google used a neural network to rapidly refine the criteria required, so that the technology can work without the need for enormous amounts of computing power.

Beyond the ImageNet large-scale visual recognition challenge, Google hopes that its research and methods will help with Google products in the future. "The progress is directly transferable to Google products such as photo search, image search, YouTube, self-driving cars, and any place where it is useful to understand what is in an image as well as where things are," said Christian Szegedy, one of the software engineers who worked on the project.

One day soon, you'll be able to search for a picture of a dog in a hat and the result will even tell you what type of hat it's wearing. Oh yes.