Face Search (1:N Identification)
Face Search identifies the most similar faces from a database by comparing embeddings and returning ranked match results.
Enrollment Response Schema
Face Enrollment Response - FaceEnrollmentResult
embeddings
- Type:
Array<number> - High-dimensional face embedding vector (e.g., 512 or 2622 dimensions).
- This vector uniquely represents the face and must be stored in a database for search operations.
Enrollment Example Response
Listing : Face Enrollment Example
{
"embeddings": [
-0.0234, 0.1245, -0.0876, 0.2341, ..., 0.0542
]
}Listing : Face Search with Matches Example
[
{
"id": "person_12345",
"distance": 0.1523,
"confidence": 0.9214
},
{
"id": "person_67890",
"distance": 0.3421,
"confidence": 0.7856
}
]Listing : Face Search with No Matches Example
[]