Google announced the launch of the Cloud Vision application programming interface, a tool developers can use to add image recognition to their own applications. The API is available in limited preview through the Google Cloud Platform, product manager Ram Ramanathan wrote.
This is a long time coming, if you ask me. Google has built up very strong technology in the area of image recognition, which can be applied to many domains, including optical character recognition (OCR), face detection, and object recognition.
“Advances in machine learning, powered by platforms like TensorFlow, have enabled models that can learn and predict the content of an image,” Ramanathan wrote. “Our limited preview of Cloud Vision API encapsulates these sophisticated models as an easy-to-use REST API.”
Even if it’s unsurprising, it’s a smart thing for Google to add in its public cloud service portfolio as the company challenges market leader Amazon Web Services, as well as Microsoft, which fields Azure and has image recognition technologies via Project Oxford.
Of course, developers can call on open-source technologies if they want to build their own, or even look to startups (like the recently funded Osaro) to find more easy-to-use image recognition services. But like other cloud providers, Google wants to give you a bunch of tools inside of one public cloud.
In terms of externalizing its machine-learning savvy, Google has long offered the Prediction API, but it’s limited in capability, with no support for images. That’s why today’s announcement is a big one for Google.