GTech Innovators

Overall, the project showcases how AWS AI services can be combined to build scalable applications for speech recognition, voice assistants, transcription services, accessibility tools, and automated content processing.

code snippet:

Steps to Build the Project:
• Step 1: Set Up an AWS Account
• Step 2: Create two S3 Buckets (Source S3 Bucket Name: amc-polly-source-bucket, Destination S3 Bucket Name: amc-polly-destination-bucket)
• Step 3: Create an IAM Policy (IAM Policy Name: amc-polly-lambda-policy)
Policy Defination:

- {
"Version": "2012-10-17",
"Statement": [
{
"Effect": "Allow",
"Action": [
"s3:GetObject",
"s3:PutObject"
],
"Resource": [
"arn:aws:s3:::amc-polly-source-bucket/*",
"arn:aws:s3:::amc-polly-destination-bucket/*"
]
},
{
"Effect": "Allow",
"Action": [
"polly:SynthesizeSpeech"
],
"Resource": "*"
}
]
}

• Step 4: Create an IAM Role (IAM Role Name: amc-polly-lambda-role) and attach amc-polly-lambda-policy and AWSLambdaBasicExecutionRole Policies
• Step 5: Create and Configure the Lambda Function (Lambda Function Name: TextToSpeechFunction)
◦ Set the runtime to Python 3.8.
◦ Set the execution role with necessary permissions for S3 and Polly. (Step 4)
◦ Add Environment Variables (SOURCE_BUCKET: Name of your source S3 bucket and DESTINATION_BUCKET: Name of your destination S3 bucket.
• Step 6: Configure S3 Event Notification
◦ Set up an event notification in the source S3 bucket to trigger the Lambda function on new object creation events with the .txt suffix.
• Step 7: Write Lambda Function Code
• Step 8: Test the System

Code:TextToSpeechFunction.py

import boto3
import json
import os
import logging

# Set up logging
logger = logging.getLogger()
logger.setLevel(logging.INFO)

def lambda_handler(event, context):
# Initialize S3 and Polly clients
s3 = boto3.client('s3')
polly = boto3.client('polly')

# Get the bucket names from environment variables
source_bucket = os.environ['SOURCE_BUCKET']
destination_bucket = os.environ['DESTINATION_BUCKET']

# Get the object key from the event
text_file_key = event['Records'][0]['s3']['object']['key']
audio_key = text_file_key.replace('.txt', '.mp3')

try:
# Retrieve text from the source S3 bucket
logger.info(f"Retrieving text file from bucket: {source_bucket}, key: {text_file_key}")
text_file = s3.get_object(Bucket=source_bucket, Key=text_file_key)
text = text_file['Body'].read().decode('utf-8')

# Send text to Polly
logger.info(f"Sending text to Polly for synthesis")
response = polly.synthesize_speech(
Text=text,
OutputFormat='mp3',
VoiceId='Joanna' # Choose the voice you prefer
)

# Save the audio file to the destination S3 bucket
if 'AudioStream' in response:
temp_audio_path = '/tmp/audio.mp3'
with open(temp_audio_path, 'wb') as file:
file.write(response['AudioStream'].read())

logger.info(f"Uploading audio file to bucket: {destination_bucket}, key: {audio_key}")
s3.upload_file(temp_audio_path, destination_bucket, audio_key)

logger.info(f"Text-to-Speech conversion completed successfully for file: {text_file_key}")

return {
'statusCode': 200,
'body': json.dumps('Text-to-Speech conversion completed successfully!')
}

except Exception as e:
logger.error(f"Error processing file {text_file_key} from bucket {source_bucket}: {str(e)}")
return {
'statusCode': 500,
'body': json.dumps('An error occurred during the Text-to-Speech conversion.')
}