Data Cleaning:
We manually processed the total number of photos in the database (>5000 images) to remove blurry photos, shots of sky or water, and duplicates. We cropped each photo in the condensed dataset (n=2267) to focus on the seal faces to minimize the amount of time the software takes to identify faces. Once photos have been cropped, they are ready to be viewed in the graphical user interface and analyzed with the face detection software (steps outlined in Figure 2 ).