The API can detect objects like boats and buildings, but will also let you detect emotion on faces. If someone in an image is happy on a boat, the Cloud Vision API will know.
The API can also help developers censor images. If you wanted to create a very PG13 version of Snapchat, for instance, it could help you filter out things kids just shouldn’t see.
Additionally, it can read text in photos, and supports a “broad set” of languages.
Use of the API is free if you keep it under 1,000 images per month, but scales as you access it more frequently.
The API is in a public beta period, where Google is limiting use to 20 million images monthly. That’s fine for smaller projects, and should even be enough for A/B testing with larger services.
➤ Vision API [Google]