First announced in December, the developer toolset can be used to add machine learning and image recognition to applications. Ultimately, the API could be applied to a range of devices -- from robots to appliances -- giving them the ability to see and understand the context of images.
In other words, a lot of future gadgets may at some point have the ability to identify your face when you walk into a room and react accordingly based on your expression.
With the Cloud Vision API, images are classified into thousands of categories in order to detect faces with associated emotions and to recognize printed words in various languages. The REST API can analyze images stored anywhere, or integrate with images stored on Google Cloud. Google said it has added additional color recognition capabilities in the beta release.
The tech giant has also announced pricing for the API, set to become effective March 1. For instance, applying Label Detection on an image could cost as little as $2 per 1,000 images, whereas Optical Character Recognition runs around 60 cents for 1,000 images.
Users will have a quota of 20 million images per month during the beta timeframe. Google noted that because of the limitation, Cloud Vision should not be used for real-time mission critical applications.