Remote Dictation Service

Description

The Remote Dictation Service integration of Invox Dictation is suitable for clients with non-Windows operating systems, such as Linux or macOS, and has certain limitations in the control of audio and microphones compared to the Local Dictation Service.

It provides the tools necessary to integrate Invox Dictation speech recognition and all its functionalities into a web environment.

The API provides from the most basic and essential functions to be able to log in and dictate in simple steps, to advanced functions to control dictation, obtain session information or even customise the behaviour of Invox Dictation components. All the low-level details related to speech recognition are hidden from the integrator.

The Invox Dictation speech recognition engine is deployed on a server, so no additional installations are required on the customer's equipment.

Integration architecture

Integration architecture

Deployment on clients

No installation on clients is required to deploy the integration. The client web application connects to the Remote Dictation Service via a WebSocket connection.

If the service does not include a valid certificate, it is necessary to include the connection exception on each client machine or to include a valid certificate.

Integration into the external application

To include Invox Dictation in your web application, import the library and execute the essential functions.

Steps to perform

  1. Import the Invox Dictation library into your web application.
    The libs/invox.min.js file contains all the necessary functionality. The following HTML block shows how to import it:
    <head>
    <script type="text/javascript" src="libs/invox.min.js" charset="UTF-8"></script>
</head>
    The Opus library libs/opus/ is mandatory for the Remote Dictation Service. Invox Dictation will automatically reference the libs/opus/ directory — keep the same folder structure:
    |-invox.min.js
|-opus/
  2. Enable MIME type wasm on the dictation server.
    This MIME type is necessary to send audio from the browser with HTML5.
    • Open Internet Information Services Manager (IIS).
    • Open MIME Types.
    • Right-click, select Add... and set:
      FieldValue
      File name extension.wasm
      MIME typeapplication/wasm
  3. Start integrating Invox Dictation into your application.
    You can start with the Quick Start guide or use the Quick Integrations to accelerate deployment.

Remarks

Invox Dictation will attempt to connect via the WebSocket protocol using the browser's native methods. It will be necessary to install the certificate of the dictation service in order to use the audio sending functionality.

Logging

See the Remote Dictation Service Logs section for details.

Table of Contents