Google Gears Up To Sell AI Vision
Monday, 07 December 2015

Google just announced a limited preview of its Cloud Vision API and moved a step closer to making money out of its AI efforts.

There isn't a huge amount of information about Google's Cloud Vision API. There are lists of what it promises to do and some idea of the nature of the API, but no in-depth technical details. 


It seems that it provides a REST API that allows you to upload images and get back results of processing. It is suggested that in the future you will be able to work with images stored in Google Cloud Storage - which makes sense. 

The list of processing options given is:

  • Label Detection

Detect broad sets of categories within an image, ranging from modes of transportation to animals.

  • Explicit Content Detection

Detect explicit content like adult content or violent content within an image.

  • Logo Detection

Detect popular product logos within an image.

  • Landmark Detection

Detect popular natural and man made structures within an image.

  • Optical Character Recognition

Detect and extract text within an image, with support for a broad range of languages, along with support for automatic language identification.

  • Face Detection

Detect multiple faces within an image, along with the associated key facial attributes like emotional state or wearing headwear.
Facial Recognition is not supported.


Essentially what we have are variations on the basic image labeling process but with different classes of label. The API includes the ability to ask for different types of label to be returned.   




You may be wondering, what is the underlying technology? It is almost certainly some variant on Google's convolutional neural network implemented using the multi-machine version of TensorFlow,  but this is speculation and not really relevant to actually using the API. 

As a demonstration Google hooked up an off-the-shelf. Raspberry Pi-based robot platform, complete with video camera, and sent it out to look at things and identify them by calling the API:



It is interesting to note Googl's emphasis that Facial Recognition is not supported. This seems to be a standard line not to be crossed for AI services. Microsoft's Project Oxford, for example, does face detection but not recognition, such is the care needed to avoid privacy issues. Yes, a good face recognizer could be used to track people, but it could also provide many useful services, including tracking people where the people in question wanted to be tracked. It seems that if you want to do anything like this you need to create your own service from scratch.






You can apply for a place on the preview. All you have to do is explain what you plan to do with the API and wait. 

What about pricing? As Google puts it:

"There is no cost for usage of the service during the Limited Preview phase. We will introduce pricing in future phases."

 So there is money to be made in AI after all.

More Information

Cloud Vision API

Related Articles

TensorFlow - Googles Open Source AI And Computation Engine

Emotion Detection Using Project Oxford

RankBrain - AI Comes To Google Search

Project Oxford Extends Reach

The Flaw In Every Neural Network Just Got A Little Worse

Google Files AI Patents

Facebook Shares Deep Learning Tools



To be informed about new articles on I Programmer, sign up for our weekly newsletter, subscribe to the RSS feed and follow us on, Twitter, FacebookGoogle+ or Linkedin




Commemorating Gordon Bell

Gordon Bell, the pioneering computer engineer who was responsible for the most successful machines of the mini computer era and co-founded the Computer History Museum, died on May 17, 2024 at the age  [ ... ]

Deno Adds Support For Private npm Registries

Deno 1.44 has been released with support for private npm registries and for gRPC connections.

More News


C book



or email your comment to:

Last Updated ( Monday, 07 December 2015 )