Amazon Polly Voice Synthesis Application

A simple web application that converts text to natural-sounding speech using Amazon Polly. Users can select different voices and enter custom text to be synthesized into audio that plays directly in the browser.

Architecture

How It Works

1. The React frontend (hosted in AWS Amplify) allows users to select a voice and enter text
2. When the user clicks "Speak", a POST request is sent to an Amazon API Gateway
3. The Amazon API Gateway forwards the request to a backend an AWS Lambda function
4. The AWS Lambda function calls Amazon Polly to synthesize speech
5. The audio is returned as base64-encoded data
6. The React frontend decodes the audio and plays it in the browser

Project Structure

my-polly-app/
├── amplify/                      # Amplify backend configuration
│   ├── backend/
│   │   ├── api/                  # API Gateway configuration
│   │   │   └── pollyApi/
│   │   └── function/             # Lambda function
│   │       └── pollyFunction/
│   │           └── src/
│   │               └── index.py  # Python Lambda handler

│   └── team-provider-info.json   # Environment configuration
├── public/                       # Static assets
├── src/                          # React source code
│   ├── App.js                    # Main application component
│   ├── App.css                   # Application styles
│   ├── index.js                  # Application entry point
│   └── aws-exports.js            # AWS configuration (auto-generated)
├── package.json                  # NPM dependencies
└── README.md                     # This file

Features

• Natural-sounding speech using Amazon Polly's neural voices
• Multiple voice options including male and female voices
• Custom text input for synthesizing any text
• Real-time audio playback directly in the browser
• Responsive design that works on desktop and mobile devices
• Loading indicators for better user experience

Technologies Used

• Frontend:

  • AWS Amplify with JavaScript library
  • React framework
  • HTML5 Audio API

• Backend:

  • Amazon API Gateway
  • AWS Lambda (Python runtime)
  • Amazon Polly
  • AWS IAM for security

Prerequisites

• Node.js and npm installed on your development machine
• AWS account with appropriate permissions
• AWS CLI installed and configured with access credentials
• AWS Amplify CLI installed (npm install -g @aws-amplify/cli)
• AWS IAM user with the following permissions:
  • IAM (for creating roles and policies)
  • Lambda
  • API Gateway
  • CloudFormation
  • S3
  • Amazon Polly

Setting Up AWS Credentials

1. Create an IAM User:

   • Sign in to the AWS Management Console
   • Navigate to IAM service
   • Create a new user with programmatic access
   • Attach policies for the required services (AdministratorAccess for simplicity, or more granular permissions)

2. Configure AWS CLI:

   aws configure

   Enter your:

   • AWS Access Key ID
   • AWS Secret Access Key
   • Default region
   • Default output format (json)

3. Configure Amplify CLI:

   amplify configure

   Follow the prompts to set up an Amplify IAM user (you can use the same user created above).

Setup and Deployment

Installation

1. Clone the repository:

   git clone https://github.com/MardiantoS/vocalize-ai-polly.git
   cd vocalize-ai-polly

2. Install dependencies:

   npm install

3. Initialize Amplify (if not already initialized):

   amplify init

   Follow the prompts to configure your project.

Local Development

The backend: API Gateway, Lambda function, and related IAM permission needs to be created first:

amplify push    # Deploy backend resources

Run the application locally:

npm start

The application will be available at http://localhost:3000

Deployment

Deploy the application to AWS:

amplify publish # Deploy frontend and backend

Key Implementation Details

The application consists of three main components:

1. React Frontend: Handles user interface, API calls, and audio playback
2. API Gateway: Provides HTTP endpoints for the frontend to communicate with the backend
3. Lambda Function: Processes requests and interacts with Amazon Polly

The Lambda function requires permissions to access Amazon Polly, which are configured through IAM roles created by Amplify.

Security Considerations

• The Lambda function has IAM permissions limited to only the required Polly actions
• API Gateway handles request validation
• CORS headers are properly configured to restrict access if needed
• Consider implementing authentication for production use

Limitations

• Audio playback requires an internet connection
• There are usage limits for Amazon Polly based on your AWS account
• Some browsers may have autoplay restrictions that require user interaction

Troubleshooting

• Audio not playing: Check browser console for errors; some browsers require user interaction before playing audio
• API errors: Verify that your Lambda function has proper permissions to access Polly
• Deployment issues: Check CloudFormation logs in the AWS Console for detailed error messages

Future Enhancements

• Add support for SSML (Speech Synthesis Markup Language)
• Implement audio download functionality
• Add language selection for multilingual support
• Integrate with other AWS services like Amazon Translate
• Add user authentication to track usage

License

This project is licensed under the MIT License - see the LICENSE file for details.

Citation

If you use this code or algorithm in your work, please cite it as follows:

This project uses vocalize-ai-polly by MardiantoS.

Acknowledgments

AWS team for the comprehensive services used in this project

---

This project was created by the author as a demonstration of AWS Amplify and Amazon Polly integration.