Google Photos has the ability to recognise the contents contained within images to help users sort and search through photos based on keyword. Now developers can embed that image recognition technology into their own apps with Google’s Cloud Vision API. The company has just released the API in limited preview on its Google Cloud Platform.
The existence of Cloud Vision is party thanks to advanced machine learning platforms like TensorFlow, which Google open sourced last month. Cloud Vision API has a number of uses for developers such as build metadata on their image catalogue, moderate offensive content.
The API can pick out the focal point of an image, read text from photos and find out what language its in, identify landmarks as well as detect faces and their conveyed emotions. Google emphasised that it does not support facial recognition and doesn’t store any facial detection information on its servers.
According to Google:
You can currently call the API by embedding an image as part of the request. In future phases, we will add support for integrating with Google Cloud Storage. The Vision API enables you to request one or more annotation types per image.
It should be noted that Google’s competitors Microsoft and Amazon both already have image recognition technologies available on their respective public cloud platforms.
If you’re interested in joining the limited preview, you can sign up by filling out this form from Google and registering your interest.