| Descriptor |
Set of unique features received from the sample. It is used for matching of faces. Descriptor is not considered personal data. |
| Cardinality |
The number of unique values a field can have. A low cardinality means a small number of unique values (e.g., 2-10), while a high cardinality means a large number (thousands or millions). |
| Index |
A tree-like data structure in Matcher, built using the specified projection fields. An index determines which fields can be used to filter data during searches and organizes data for quick access. |
| Projection |
A specialized dataset prepared for matching. Contains a filtered subset of records with a specific set of attributes. |
| Matching |
The process of comparing descriptor with some descriptor batch that results in similarity scores. Its purpose is to search most similar descriptors to a given one over some user-provided set of descriptors. |
| Sample |
An image containing a face or body that meets VisionLabs and other standards. |
| Targets |
A list of event attributes that are stored in the projection along with the descriptor. These determine which metadata will be available for filtering and analysis during search. |