Text Independent - Identify Single Speaker

識別單一說話者配置檔
識別在候選配置檔清單中在輸入音訊中說話的人員。

限制:

  • 音訊輸入長度下限為 1秒

  • 音訊輸入長度上限為 120秒

  • 候選演講者計數下限為 1

  • 候選說話者計數上限為 50

  • 最小有效語音長度 (排除無聲和其他非語音畫面) 為 4 秒 此限制可藉由將 “ignoreMinLength” 設定為 true 來停用。

  • (SNR) 的音訊號與雜訊比率下限為 2dB

POST {endpoint}/speaker-recognition/identification/text-independent/profiles:identifySingleSpeaker?api-version=2021-09-05&profileIds={profileIds}
POST {endpoint}/speaker-recognition/identification/text-independent/profiles:identifySingleSpeaker?api-version=2021-09-05&profileIds={profileIds}&ignoreMinLength={ignoreMinLength}

URI 參數

名稱 位於 必要 類型 Description
endpoint
path True

string

支援的認知服務端點 (通訊協定和主機名,例如: https://westus.api.cognitive.microsoft.com) 。

api-version
query True

string

指定用於這個要求的作業版本。

profileIds
query True

string[]

以逗號分隔的配置檔標識碼。 支持的數目上限為50個標識碼。

ignoreMinLength
query

boolean

如果為 true,則會略過識別所需的最小語音數量。 預設值為 false。

要求標頭

Media Types: "audio/wav; codecs=audio/pcm"

名稱 必要 類型 Description
Ocp-Apim-Subscription-Key True

string

要求本文

Media Types: "audio/wav; codecs=audio/pcm"

名稱 類型 Description
audioData

object

二進位音訊檔案。 支援的格式為音訊/wav;codecs=audio/pcm。 支援最多 5MB 的音訊。

回應

名稱 類型 Description
200 OK

IdentifiedSingleSpeakerInfo

確定

Other Status Codes

SpeakerErrorInfo

失敗

Headers

x-ms-error-code: string

安全性

Ocp-Apim-Subscription-Key

Type: apiKey
In: header

範例

Successful Query

Sample Request

POST https://westus.api.cognitive.microsoft.com/speaker-recognition/identification/text-independent/profiles:identifySingleSpeaker?api-version=2021-09-05&profileIds=3669fa29-1bf3-45ad-beea-6b348d058d7e,111f427c-3791-468f-b709-fcef7660fff9,0e196cd9-32d5-4883-8631-54a0e7c7cb3d,0e196cd9-32d5-4883-8631-54a0e7c7cb3d,726e57d9-04e0-4214-b482-7f786fa83560,f95189fd-1bf5-4485-9c2e-e5897e0c98ca


"{binary file date}"

Sample Response

Content-Type: application/json
{
  "identifiedProfile": {
    "profileId": "111f427c-3791-468f-b709-fcef7660fff9",
    "score": 0.63
  },
  "profilesRanking": [
    {
      "profileId": "111f427c-3791-468f-b709-fcef7660fff9",
      "score": 0.63
    },
    {
      "profileId": "3669fa29-1bf3-45ad-beea-6b348d058d7e",
      "score": 0.49
    },
    {
      "profileId": "0e196cd9-32d5-4883-8631-54a0e7c7cb3d",
      "score": 0.4
    },
    {
      "profileId": "726e57d9-04e0-4214-b482-7f786fa83560",
      "score": 0.1
    },
    {
      "profileId": "f95189fd-1bf5-4485-9c2e-e5897e0c98ca",
      "score": 0.03
    }
  ]
}
Content-Type: application/json
x-ms-error-code: Error Code
{
  "error": {
    "code": "Error Code",
    "message": "Erro Messae"
  }
}

定義

名稱 Description
Error
IdentifiedSingleSpeakerInfo
IdentifyInfo

識別的說話者資訊

SpeakerErrorInfo

說話者錯誤訊息

Error

名稱 類型 Description
code

string

message

string

IdentifiedSingleSpeakerInfo

名稱 類型 Description
identifiedProfile

IdentifyInfo

物件,其中包含已識別配置檔的數據。

profilesRanking

IdentifyInfo[]

物件,其中包含前 5 個設定檔的數據, (包括已識別的配置檔) 依分數排序的遞減順序。

IdentifyInfo

識別的說話者資訊

名稱 類型 Description
profileId

string

已識別配置文件的識別碼。 如果未將候選項目識別為正確的說話者,此值會設定為空的 GUID。

score

number

浮點數,指出輸入音訊與目標語音列印之間的相似度。 此數字必須介於 0 到 1 之間。 較高的數位表示較高的相似度。

SpeakerErrorInfo

說話者錯誤訊息

名稱 類型 Description
error

Error