In order to take advantage of these types of metadata, I need to split them up into smaller categories. For example, the Face Detection metadata returns the following attributes:
Ideally, each key (joy, sorrow, etc...) has its own collection, so that graph traversal is easier in cases where a User searches for a single emotion in a picture, such as anger.
Both
FACE_DETECTION
andSAFE_SEARCH_DETECTION
use the enum values for their keys:In order to take advantage of these types of metadata, I need to split them up into smaller categories. For example, the Face Detection metadata returns the following attributes:
Ideally, each key (
joy
,sorrow
, etc...) has its own collection, so that graph traversal is easier in cases where a User searches for a single emotion in a picture, such asanger
.The same applies for the SafeSearch model