Malayalam to English Automatic Subtitle Generation

Tool to transcribe Malayalam audios into English and format it as a .srt file with the proper timings. It uses OpenAI's 'Whisper' model to translate from Malayalam to English.

I created this script as I need this tool to create subtitles for sermons at my church. I hope this tool can be useful to you as well for transcribing Malayalam audios to English. If you liked this please consider leaving a ⭐, and share it, so others can also find it helpful.

Run in Colab

The tool can be used either locally or on web hosted services such as Colab. I have written one for you which you can use to generate subtitles.

The colab file can be found here

Run Locally

Setup (First Run)

1. Install OpenAI's Whisper on your local computer via Python's package manager pip.

   pip install -U openai-whisper

2. Whisper requires you to have ffmpeg installed on your system, which is available from most package managers.

   # on Ubuntu or Debian
   sudo apt update && sudo apt install ffmpeg
   
   # on Arch Linux
   sudo pacman -S ffmpeg
   
   # on MacOS using Homebrew (https://brew.sh/)
   brew install ffmpeg
   
   # on Windows using Chocolatey (https://chocolatey.org/)
   choco install ffmpeg
   
   # on Windows using Scoop (https://scoop.sh/)
   scoop install ffmpeg

3. You will require setuptools-rust as well (to avoid unecssary errors).

   pip install setuptools-rust
   

Command-Line Usage

1. Clone the git repo to your system via command-line or by downloading the .zip file (at the top of this page).

   git clone https://github.com/Roshan-Thomas/malayalam_english_subtitle_generator.git
   

2. Change directory to newly downloaded folder

   cd malayalam_english_subtitle_generator

3. Run the following command to generate subtitles for a sample audio file 'sample.wav'.

   python translate_message_to_srt.py -a "sample.wav" 
   
   # Or, you can run the following
   python translate_message_to_srt.py --audio "sample.wav"

Optional Commands

1. To specify a custom Whisper model, use the -m or --model argument

   python translate_message_to_srt.py -a "sample.wav" -m large
   
   # Or you can use the following
   python translate_message_to_srt.py -a "sample.wav" --model large

2. To specify a custom save location for the generated .srt files, use the -s or --save argument

   python translate_message_to_srt.py -a "sample.wav" -s save_location_file_path
   
   # Or, you can use the following
   python translate_message_to_srt.py -a "sample.wav" --save save_location_file_path

3. To specify a custom load location for your audio (.wav or .mp3) file, use the -l or --load argument

   python translate_message_to_srt.py -a "sample.wav" -l load_location_file_path
   
   # Or, you can use the following
   python translate_message_to_srt.py -a "sample.wav" --load load_location_file_path

4. Command which uses all the above optional arguments

   python translate_message_to_srt.py -a "sample.wav" -l audios -s srt_files -m large