Developer Center / Getting Started / Hands-on / ...
Create an Audio Transcript
with Amazon Transcribe
In this step-by-step tutorial, you will learn how to use Amazon Transcribe to create a text transcript of a recorded audio file using the AWS Management Console. Amazon Transcribe is an automatic speech recognition (ASR) service that makes it easy for developers to add speech-to-text capability to their applications. Using the Amazon Transcribe API, you can analyze audio files stored in Amazon Simple Storage Service (S3) and have the service return a text file of the transcribed speech.
As a developer, creating transcriptions of customer service calls or generating subtitles on audio and video content are common challenges requiring speech-to-text capabilities. This challenge could be solved by building your own machine learning models from scratch. However, this option is time-intensive, expensive, and requires machine learning expertise. Instead of taking the difficult route, you can use Amazon Transcribe, a pre-trained and fully-managed service, which provides fast and high-quality transcriptions.
In this tutorial you will download a sample audio file then upload it to a S3 bucket that you will create. Then you will use Amazon Transcribe to create a transcript from the sample audio clip using the AWS Management Console.
This tutorial is a demo of the functionality that is available when using the AWS CLI or the Transcribe API. For production or proof of concept implementations, we recommend using these programmatic interfaces rather than the Amazon Transcribe Console.
This tutorial requires an AWS account
It’s easy to get started with the Amazon Transcribe Free Tier. Transcribe up to 60 minutes of audio each month - free for the first 12 months your AWS account has been open.
Step 1. Create a S3 bucket and upload sample audio file
In this step you will download a sample audio file, create a S3 bucket, then upload the sample file to the S3 bucket. Transcribe accesses audio and video files for transcription exclusively from S3 buckets.
a. When you click here, you will download the sample audio file to transcribe later in the tutorial. The file name is transcribe-sample.mp3.
b. When you click here, the AWS Management Console will open in a new browser window, so you can keep this step-by-step guide open. When the screen loads, enter your user name and password to get started. Using the Region drop down, select a Region that has Amazon Transcribe.
At the time of writing, Amazon Transcribe is available in US East (N. Virginia), US East (Ohio), US West (Oregon), Asia Pacific (Sydney), Canada (Central), and EU (Ireland) regions.
e. Enter a unique bucket name. Bucket names must be unique across all existing bucket names in Amazon S3. There are a number of other restrictions on S3 bucket names as well. Then select a Region to create your bucket in.
Select Next.
f. You have many useful options for your S3 bucket including Versioning, Server Access Logging, Tags, Object-level Logging and Default Encryption. We won't enable these features for this tutorial.
Select Next.
Step 2. Create transcription job
In this step you will create and run a transcription job using the Amazon Transcribe console.
d. On the Create transcription job page, in the Name field, type sample-transcription-job.
Leave the default Language as English.
In the Input file location on S3 field, paste the link to the sample file in your S3 bucket. The link to your sample file will be different than the one shown in the screenshot to the right.
Leave the default Format of mp3.
You can use the Custom vocabulary feature to help Amazon Translate recognize words and phrases that are specific to your application, such as a non-English name like Etienne. You won't use this feature in this tutorial.
e. Leave the default Data location as Amazon default.
if you enable Channel identification, Amazon Transcribe is able to process audio where each speaker is recorded on different channels and produce a single transcript annotated with channel labels.
If you enable Speaker identification, Amazon Transcribe recognizes when the speaker changes and attributes the transcribed text appropriately.
Neither of these options is used in this tutorial.
Select Create to start your transcription job.
Step 3. Review transcription results
In this step, you will learn how to check on the progress and review the results of your transcription job.
a. After you click the Create button, you will be taken to the Transcription jobs screen. It will show the status of sample-transcription-job. The status can be In progress, Complete, or Failed.
When the status is Complete, click on the sample-transcription-job link in the Name column to view the transcription results.
Step 4. Terminate resources
In this step you will delete the sample file from your S3 bucket to avoid unnecessary charges.
Congratulations!
As you have seen in this tutorial, Amazon Transcribe enables voice to text at scale. Use Amazon Transcribe for a wide range of audio or videos files, such as customer service calls, business meetings, broadcast TV, and on-demand videos.
Recommended Next:
Learn to Translate Text at Scale
Find out how you can translate transcribed texts into other language using Amazon Translate in this 10-minute tutorial.
Get Started with the Transcribe API and CLI
Learn how to use the Amazon Transcribe API and CLI in this tutorial in documentation.
Explore Amazon Comprehend
Learn how you can analyze sentiment in text using Amazon Comprehend in this 10-minute tutorial.