Google just released a new machine vision tool for developers to help them build applications that can “understand” the content of images. In a blog post, the company announced the beta version of its Cloud Vision application programming interface (API) on Wednesday. Currently, only select developers get the chance to try out the six tools built to help application handle photos more easily.
The Safe Search Detection tool could be especially helpful for companies that build apps to share crowdsourced images. We all know just because an app says not to post certain things, it doesn't mean they won't show up anyway. The Safe Search Detection tool gives developers access to the powerful technology used by Google’s SafeSearch to block inappropriate content.
Another feature, Facial Detection, can identify a lot more than just a face. The feature detects facial features and their placement, and the emotion in the face such as “joy and sorrow.” And Google promises to not use the tool for facial recognition, or store any of the information on their servers.
Landmark Detection allows an app to identify a photo of the Eiffel Tower as the Eiffel Tower, not just some random tower, and locate where on a map it is. It also works for popular natural landmarks. Entity Detection picks up on the dominant thing in a photo and identifies it. A photo from the farmer’s market might be labeled “produce”.
If you're a developer and want access to the Cloud Vision API now, fill out the survey and explain what you plan to do with the tools.