multipaz-vision/org.multipaz.facematch

Package-level declarations

Types

data class FaceEmbedding(val embedding: FloatArray)

The normalized multidimensional face embeddings vector uniquely describing the face features per LiteRT model (e.g. FaceNet).

FaceMatchLiteRtModel

data class FaceMatchLiteRtModel(val modelData: ByteString, val imageSquareSize: Int, val embeddingsArraySize: Int)

Data class for the the platform independent LiteRT model handling.

Functions

getFaceEmbeddings

actual fun getFaceEmbeddings(image: ImageBitmap, model: FaceMatchLiteRtModel): FaceEmbedding?

expect fun getFaceEmbeddings(image: ImageBitmap, model: FaceMatchLiteRtModel): FaceEmbedding?

Extract face embeddings in a normalized values array. The method supposed to be called from the Camera composable onFrameCaptured callback which is run on IO thread.

actual fun getFaceEmbeddings(image: <Error class: unknown class>, model: <Error class: unknown class>): <Error class: unknown class>?