Transcribe

Category: AWS

|

Runtime: 1.0+

|

Available: Free

Overview

Transcribe node by Kumologica integrates with Amazon Transcribe (opens new window) service for converting a speech audio file to text.

Setup

Prerequisites

For Transcribe node to work properly, you need to have a valid subscription of AWS cloud account. Ensure to have an Access credentials (opens new window) generated for the node to connect with the Transcribe service associated with the account.

Installation

To install the node from designer:

1. Go to Add more nodes option on the palette section.
   
2. Check for the Amazon Transcribe node and click install.
   
3. Once installed, the designer is required to be restarted.
   

To install the node from CLI:

1. Go to your project workspace where you can see your package.json file.
   
2. Run the below give npm command.
   npm i @kumologica/kumologica-contrib-awstranscribe

Technical Details

Properties

• Job Name (Required)
  Name of the job that transcribe speech to text.
• Media S3 Bucket (Optional)
  Name of the S3 bucket where the media file (wav or mp3 etc) is stored.
• Media Path (Required)
  Path to the media file (wav or mp3 etc) stored in the S3 bucket. Path must include the media filename with extension. Eg: /mymediafolder/hello.wav
• Output S3 BucketName (Optional)
  Amazon S3 bucket name where the result to be stored.
• Language Code (Optional)
  Language code associated with the media file for proper conversion.
• Media Format (Optional)
  Media format associated with the audio file. Supports MP3, MP4, WAV etc.

Supported Operations

• startTranscriptionJob
• Delete Transcribe Job
• Get Transcribe Job Details

Throws

Transcribe node throws the following exception.

• AWSTranscribeErrorResponse
• AWSClientError

Returns

Payload will be overwritten by the response from the node. It will be having the message object based on the operation executed by node. Payload can be accessed on the subsequent node using msg.payload.