Class: ReplicateFaceEmbeddingService
Defined in: packages/agentos/src/io/media/images/face/ReplicateFaceEmbeddingService.ts:61
Extracts face embeddings via the Replicate API using InsightFace.
Sends a prediction request with the provided image URL, polls until the prediction completes, then parses the embedding vector from the response.
Implements
Constructors
Constructor
new ReplicateFaceEmbeddingService(
config):ReplicateFaceEmbeddingService
Defined in: packages/agentos/src/io/media/images/face/ReplicateFaceEmbeddingService.ts:66
Parameters
config
Returns
ReplicateFaceEmbeddingService
Methods
compareFaces()
compareFaces(
a,b,threshold?):FaceComparisonResult
Defined in: packages/agentos/src/io/media/images/face/ReplicateFaceEmbeddingService.ts:168
Compare two face embedding vectors using cosine similarity.
Parameters
a
number[]
First embedding vector.
b
number[]
Second embedding vector.
threshold?
number = 0.6
Minimum similarity to consider a match (default 0.6).
Returns
FaceComparisonResult
Comparison result with similarity score and match flag.
Implementation of
IFaceEmbeddingService.compareFaces
extractEmbedding()
extractEmbedding(
imageUrl):Promise<FaceEmbedding>
Defined in: packages/agentos/src/io/media/images/face/ReplicateFaceEmbeddingService.ts:81
Extract a face embedding from an image.
Parameters
imageUrl
string
Public URL or base64 data URI of the image.
Returns
Promise<FaceEmbedding>
Face embedding with 512-dim vector and optional bounding box.