Google has launched a new tool that shows how they analyse and interpret an image. Using their cloud learning and artificial intelligence algorithms it enables you to see what Google thinks the image is relevant for.

Not only is this technology very clever, the tool is pretty useful for web and SEO teams as it gives an insight in to how Google AI and machine learning tools go about understanding an image.

What is the Google AI Image Recognition Tool?

The tool is a great way to demo Google’s Cloud Vision API. Simply upload an image and it tells you how Google’s machine learning algorithm interprets it.

Try the tool

We tested the new tool with the image below and some of things Google is able to recognise in the photo is truly astounding.

 

These are seven ways Google’s image analysis tools classifies uploaded images:

  1. Faces
  2. Objects
  3. Labels
  4. Web Entities
  5. Text
  6. Properties
  7. Safe Search

Faces

The “faces” tab provides an analysis of the emotion expressed by the image. The accuracy of this result can be a little questionable – and definitely depends on the quality of the image.

However, you can clearly see that Google recognises at least two faces in our photo – and then has a good guess at the emotion on each of the faces – in this case getting it pretty spot one with a “joyful” face.

Objects

The “objects” tab shows what objects are in the image, like glasses, person, etc. This works very well and you can see Google’s AI tool identifies a male, female and a dog. (PS I thought the dog was going to be pretty hard to get given it looks like a ball of fur!)

Labels

The “labels” tab shows details about the image that Google recognises, like  people and grass but then also captures the conceptual aspects like “people in nature” and fun.

Web Entities

This shows descriptive words that are associated with the image that can be found via the web. There are a couple of things that are slightly dubious like “obedience trial” – but for most things again, it gets it pretty much spot on.  It correctly identifies that the image is also used on the Highlands End Holiday Park website – one of the images in their gallery.

The Web Entities tab is quite a useful element of the tool as it shows how Google itself is interpreting what the image means by what is published online with that image.

Properties

Properties highlights the colours that are used in the image.

Safe Search

Safe search shows how the image ranks for unsafe content. The descriptions of potentially unsafe images are as follows:

  • Adult
  • Spoof
  • Medical
  • Violence
  • Racy

Key tip

This latest tool form the Google camp is an interesting insight into how Google might possibly understand your images. It could also give you a hint if your images need further optimising.

You can try the tool for yourself by following the link below and uploading an image:-

https://cloud.google.com/vision/docs/drag-and-drop

Follow my blog with Bloglovin