Behind the Scenes: Google Explains How it Identifies Landmarks

Thanks to machine-learning algorithms, Google Photos can detect the landmarks in your photos. Third-party developers can use these powerful algorithms in their own apps, too, and Google is trying to sell them on why they should.

Google uses a multi-step process, which its developers refer to as a "landmark mining pipeline," to identify which point of interest you've caught in front of your lens. Here's how it works, according to a post on Google's engineering blog:

"If the pipeline identifies an image as a possible landmark, it is sent to the next step of the pipeline for clustering. Photos are then clustered in two ways: by geolocation data and by visual match. One cluster is mapped to a single landmark. For example, a cluster could include multiple photos of the Statue of Liberty taken at different angles, some with objects in the foreground, and some with textual metadata that helps identify the landmark. Once a cluster is identified, the pipeline generates the associated metadata for the landmark: its name, Knowledge Graph mid, and latitude / longitude coordinates. Clusters that aren't associated with a real landmark are discarded."

The secret weapon is that, thanks to all those different angles and metadata, the algorithms can identify not only the landmark, but also whether or not you've taken a photo of something similar. Did you visit the Statue of Liberty replica in Paris's River Seine? Google knows. Snap a picture of the mini Eiffel Tower at Las Vegas's Paris Hotel? Google will tell you.

Share this

Share on FacebookPlus on Google+