🎙️ English Subtitle Generator (Speech Recognition) using Whisper

🔎 About

This project focuses english speech reconition and transform it into subtitles (.srt) using the Whisper (medium.en) model.

Project languages:

• en

📦 Dependencies

Name	Version
streamlit	1.34.2
srt	3.5.3
whisper	1.1.10

🖥️ Requirements

• Operating System (OS): Windows 10, Mac, Linux.
• python==3.9.9 and pytorch==1.10.1
• Integrated Development Environment (IDE): VSCode.
• Web Browser: Google Chrome, Microsoft Edge, Firefox, Safari.

⬇️ Installation

Make a directory

mkdir img_segmentation_app

cd img_segmentation_app

Create and activate environment

python -m venv venv

venv\Scripts\activate 

Install dependencies

pip install openai-whisper streamlit srt ffmpeg

setup FFmpeg:

1. Download and extract FFmpeg in C Drive: https://ffmpeg.org/download.html
2. Create new terminal
3. run:

   $env:PATH = "C:\ffmpeg\bin;" + $env:PATH
   

4. activate venv again:

   venv\Scripts\activate 
   

Run App.

streamlit run app.py

🥼 Author(s) / Contributor(s)

• Wicaksono Hanif Supriyanto

📚 References

• Radford, A., Kim, J. W., Xu, T., Brockman, G., McLeavey, C., & Sutskever, I. (2022). Robust speech recognition via large-scale weak supervision. arXiv. https://arxiv.org/abs/2212.04356
• Inspired by Bikin Subtitle Otomatis dengan Python (Python Tutorial) by ngodein.