kiyodori / speech_to_ae_subtitle

speech_to_ae_subtitle

Script for converting speech to subtitle.

Requirements

• Adobe After Effects

Setting

Create and launch a Docker container.

# Create and launch a Docker container
$ docker-compose up -d --build

# Run the container bash
$ docker-compose exec app bash

Create an AWS IAM user and two S3 buckets for file input and file output. Then, Set config and env files.

$ cp .aws/config.dev .aws/config
$ cp .env.dev .env

Create AWS lambda function for transcribing when wav file has uploaded to S3 input bucket.

• Set transcribe_function/lambda_function.py to AWS lambda function
• Specify an S3 input bucket for the lambda trigger
• Set the following two lambda environment variables
  • OUTPUT_BUCKET: AWS S3 output bucket
  • REGION: AWS S3 region
• Attach the following policy to lambda function
  • AmazonS3FullAccess
  • AmazonTranscribeFullAccess
  • CloudWatchLogsFullAccess

Usage

1. Convert mp4 to wav

Place the mp4 files you want to convert in the input directory. The following command will output a wav file under the audio directory.

$ poetry run python converter.py test.mp4 test.wav

2. Upload wav to Amazon S3

Create S3 bucket. Then upload the wav file to the Amazon S3 bucket.

$ poetry run python uploader.py test.wav

3. Create AE script

4. Modify texts of AE script

5. Execute AE script

6. Export to mov

Development

You can check code format.

# Lint
$ make lint

# Format
$ make format