次の方法で共有


Call Media - Recognize

通話からメディアを認識します。

POST {endpoint}/calling/callConnections/{callConnectionId}:recognize?api-version=2024-09-15

URI パラメーター

名前 / 必須 説明
callConnectionId
path True

string

呼び出し接続 ID

endpoint
path True

string

url

Azure Communication リソースのエンドポイント。

api-version
query True

string

呼び出す API のバージョン。

要求ヘッダー

名前 必須 説明
Authorization True

string

Azure Communication Services ユーザー アクセス トークン。

要求本文

名前 必須 説明
recognizeInputType True

RecognizeInputType

認識の種類を決定します。

recognizeOptions True

RecognizeOptions

認識のオプションを定義します。

interruptCallMediaOperation

boolean

set recognize が他の既存のキューに登録されている要求または現在処理中の要求に割り込むことができます。

operationCallbackUri

string

この操作の CreateCall/AnswerCall によって設定された既定のコールバック URI をオーバーライドするコールバック URI を設定します。 このセットアップはアクションごとに行われます。 これが設定されていない場合は、CreateCall/AnswerCall によって設定された既定のコールバック URI が使用されます。

operationContext

string

操作のコンテキストを識別する値。

playPrompt

PlaySource

認識のために再生するオーディオのソース。

playPrompts

PlaySource[]

認識のために再生するオーディオのソース。

応答

名前 説明
202 Accepted

サービスは認識要求を受け入れ、処理を開始します。 指定したコールバック URI で RecognizeCompleted イベントまたは RecognizeFailed イベントを受け取り、要求の状態を更新します。

Other Status Codes

CommunicationErrorResponse

エラー

セキュリティ

Authorization

Azure Communication Services ユーザー アクセス トークン。

型: apiKey
/: header

CallMedia_Recognize

要求のサンプル

POST https://contoso.communications.azure.com/calling/callConnections/18dea47f-b081-4107-9a5c-4300819d2c6c:recognize?api-version=2024-09-15

{
  "recognizeInputType": "dtmf",
  "playPrompt": {
    "kind": "file",
    "file": {
      "uri": "https://some.file.azure.com/sample.wav"
    }
  },
  "recognizeOptions": {
    "interruptPrompt": true,
    "initialSilenceTimeoutInSeconds": 5,
    "targetParticipant": {
      "kind": "communicationUser",
      "communicationUser": {
        "id": "8:acs:b9614373-fd0b-480c-8fd2-cb58b70eab9f_da7be3a9-8788-42a6-85c6-56b2cf784fce"
      }
    },
    "dtmfOptions": {
      "interToneTimeoutInSeconds": 3,
      "maxTonesToCollect": 5,
      "stopTones": [
        "pound"
      ]
    }
  },
  "operationCallbackUri": "https://app.contoso.com/callback"
}

応答のサンプル

定義

名前 説明
Choice
DtmfOptions

DTMF 認識のオプション

FileSource
PlaySource
PlaySourceType

再生ソースの種類を定義します

RecognizeInputType

認識の種類を決定します。

RecognizeOptions
RecognizeRequest
SpeechOptions

継続的な音声認識のオプション

SsmlSource
TextSource
Tone
VoiceKind

音声の種類

Choice

名前 説明
label

string

特定の選択肢の識別子

phrases

string[]

認識する語句の一覧

tone

Tone

DtmfOptions

DTMF 認識のオプション

名前 説明
interToneTimeoutInSeconds

integer

DTMF 入力が認識を停止するまで待機する時間。

maxTonesToCollect

integer

収集する DTMF トーンの最大数。

stopTones

Tone[]

認識を停止するトーンの一覧。

FileSource

名前 説明
uri

string

再生するオーディオ ファイルの URI

PlaySource

名前 説明
file

FileSource

再生に使用するファイル ソース情報を定義します

kind

PlaySourceType

再生ソースの種類を定義します

playSourceCacheId

string

関連メディアのキャッシュに使用する識別子を定義します

ssml

SsmlSource

再生に使用する ssml(Speech Synthesis Markup Language) ソース情報を定義します

text

TextSource

再生に使用するテキスト ソース情報を定義します

PlaySourceType

再生ソースの種類を定義します

名前 説明
file

string

ssml

string

text

string

RecognizeInputType

認識の種類を決定します。

名前 説明
choices

string

dtmf

string

speech

string

speechOrDtmf

string

RecognizeOptions

名前 説明
choices

Choice[]

認識の Ivr の選択肢を定義します。

dtmfOptions

DtmfOptions

DTMF の構成を定義します。

initialSilenceTimeoutInSeconds

integer

プロンプトの後に最初の入力を待機する時間 (ある場合)。

interruptPrompt

boolean

プロンプトを中断して認識を開始するかどうかを決定します。

speechLanguage

string

認識される音声言語、既定値が設定されていない場合は en-US

speechOptions

SpeechOptions

継続的な音声認識オプションを定義します。

speechRecognitionModelEndpointId

string

カスタム モデルがデプロイされたエンドポイント。

targetParticipant

CommunicationIdentifierModel

RecognizeRequest

名前 説明
interruptCallMediaOperation

boolean

set recognize が他の既存のキューに登録されている要求または現在処理中の要求に割り込むことができます。

operationCallbackUri

string

この操作の CreateCall/AnswerCall によって設定された既定のコールバック URI をオーバーライドするコールバック URI を設定します。 このセットアップはアクションごとに行われます。 これが設定されていない場合は、CreateCall/AnswerCall によって設定された既定のコールバック URI が使用されます。

operationContext

string

操作のコンテキストを識別する値。

playPrompt

PlaySource

認識のために再生するオーディオのソース。

playPrompts

PlaySource[]

認識のために再生するオーディオのソース。

recognizeInputType

RecognizeInputType

認識の種類を決定します。

recognizeOptions

RecognizeOptions

認識のオプションを定義します。

SpeechOptions

継続的な音声認識のオプション

名前 説明
endSilenceTimeoutInMs

integer

ユーザーが話しを停止し、cogservice が応答を送信するときの終了無音の長さ。

SsmlSource

名前 説明
customVoiceEndpointId

string

カスタム音声がデプロイされたエンドポイント。

ssmlText

string

再生するコグニティブ サービスの Ssml 文字列

TextSource

名前 説明
customVoiceEndpointId

string

カスタム音声がデプロイされたエンドポイント。

sourceLocale

string

再生するソース言語ロケール ここで使用可能なロケールを参照してください。

text

string

再生するコグニティブ サービスのテキスト

voiceKind

VoiceKind

音声の種類

voiceName

string

再生する音声名 ここで使用可能なテキスト読み上げ音声を参照してください。

Tone

名前 説明
a

string

asterisk

string

b

string

c

string

d

string

eight

string

five

string

four

string

nine

string

one

string

pound

string

seven

string

six

string

three

string

two

string

zero

string

VoiceKind

音声の種類

名前 説明
female

string

male

string