Skip to main content

Voiceprint Comparison: start task

POST 
/api/technology/speaker-identification-voiceprint-comparison

Start Voiceprint Comparison task for two voiceprint lists.

The voiceprint lists are provided in the voiceprints_a and voiceprints_b parameters. The size of the lists must meet the following criteria:

  • There must be at least one voiceprint in each list.
  • The size of each list must not exceed the limit for maximum length, which defaults to 50,000.
  • The product of the two list sizes must not exceed the limit for maximum voiceprint matrix size, which defaults to 1,000,000 (e.g., possible maximum list lengths include len(voiceprints_a)=1_000, len(voiceprints_b)=1_000 and len(voiceprints_a)=50_000, len(voiceprints_b)=20).

Each voiceprint is expected to be a Base64-encoded string.

How a Voiceprint Comparison request is arranged​

See Voiceprint Comparison task result documentation for details on how the result is arranged.

The request consists of two voiceprint lists arranged as columns and rows of the resulting score matrix.

As a result, all voiceprints from the columns list will be compared to all voiceprints in the rows list. The resulting scores will be filled in a matrix of (sizeof(voiceprints_a), sizeof(voiceprints_b)) dimensions.

For example, the following input:

{
  "voiceprints_a": ["vp_a_1", "vp_a_2", "vp_a_3"],
  "voiceprints_b": ["vp_b_1", "vp_b_2", "vp_b_3"]
}

encodes the following empty Voiceprint Comparison score input matrix:

voiceprints_a list
vp_a_1vp_a_2vp_a_3
voiceprints_b listvp_b_1
vp_b_2
vp_b_3

The Voiceprint Comparison technology will then fill in the empty cells with scores.

Request​

Responses​

Voiceprint Comparison task was accepted. Follow the X-Location header to poll for the task state.

Response Headers
    X-Location

    A URL the client should poll for task state and result.

    Example: /api/technology/speaker-identification-voiceprint-comparison/651f10f1-3025-41e7-ad4f-d4ae385c8b7d