Identity saving or identification happens after central embedder creates a vector (central embedding process).
This is where central embedder converts the face attributes into a vector consisted of 512 numbers. Embedding calculation and attributes prediction is performed for every detection of sufficient quality within sighting.
Each identity's central vector metadata is saved centrally on the cloud. This way we're able to recognize same identities on multiple locations due to the data centralization. Sighting central embedding vector is in the identification process.
Each time an identity shows up in front of a camera, embedder creates a new vector from a sighting. So if subject appears 10 times embedder will create a new vector each time. Central embedder does the conversion on a locally dedicated server and sends it to a cloud database. So, all vectors ever created are stored on a central cloud service.
Every vector is being compared to all existing identitie's vectors to check if any one of those has a similar value. If there is an existing vector with a similar value it joins the new one and enters the embedding vector collection that is representation of one identity. The system tracks the similarity of vectors with beforehand set thresholds. Slightly similar vectors are listed as similar identities.
If a new vector is different from any existing vector, the system creates a new identity.