AudioNote

Project Summary

AudioNote lets you transcribe your notes.

Installation

Prerequisites

AudioNote uses AWS to upload files and perform transcriptions. Before you begin...

1. Create an AWS account.

2. Create an S3 bucket to which audio files will be uploaded.

   Installation Steps

3. Obtain your AWS Access Key Id and Secret Access Key. Also take note of the bucket name in S3.

4. To test the backend locally, you must set a few environmental variables on your computer by running the following in the command line:

   export AWS_ACCESS_KEY_ID=YOUR_ACCESS_ID // Access ID from AWS
   export AWS_SECRET_ACCESS_KEY=YOUR_ACCESS_SECRET // Access secret from AWS
   export AWS_REGION=us-east-1 // AWS region where you want to run AudioNote
   export AWS_BUCKET_NAME=YOUR_BUCKET_NAME // AWS bucket where you want audio files to be stored. Must be in same region as above.
   export AUDIONOTE_TOKEN=1234 // Token AudioNote will use to authenticate requests. Include it as the 'Token' header in all requests. 

5. You can start the server by moving to the 'Backend' directory and running mvn spring-boot:run.

6. You can run the app in Android Studio.

7. You can access the backend hosted on Heroku at https://audionoteucsb.herokuapp.com.

   • Set "Token" header to 1234
   • Send a POST request with an audio file under paramater "audio" to https://audionoteucsb.herokuapp.com/transcription
   • Check the status of a job and view the results by sending a GET request to https://audionoteucsb.herokuapp.com/transcription/YOUR_JOB_ID with the job id you recieved in the previous request.

Resources

AWS Sagemaker: https://aws.amazon.com/sagemaker/

Mozilla Common Voice: https://voice.mozilla.org/en