CoralNet API documentation

The CoralNet deploy API lets you programmatically classify any image using any automatic classifier on CoralNet. You specify images using web URLs along with the row, column locations you want classified and you get the automated classifications back. Easy as pie!

The classifiers are tied to the source they were trained for and thus uses the label-set of that source. You can use any classifier from any public source or any source for which you have view/edit/admin permissions.

This API allows a two-stage workflow for automated classification:

  • Classifier curation. This is support by the current CoralNet site. Upload images -> annotate some images -> get a new classifier -> use new classifier to annotate more images faster -> get better classifier. And so on.

  • Deploy. Once satisfactory classifier performance is achieved, you can call the deploy API to automatically and programmatically classify all images from your survey. This allows for 10-100 thousands of images to be classified. (Note: if you are planning to use this for more than 100k images, please let us know before you start.)

Ready to start? Check out more details below.