AI-Youtube-Shorts-Generator

An AI-powered tool that automatically generates engaging short-form videos from longer YouTube content. Forked from SamurAIGPT/AI-Youtube-Shorts-Generator

Features

• Video Download: Given a YouTube URL, the tool downloads the video.
• Transcription: Uses Whisper to transcribe the video.
• Highlight Extraction: Utilizes OpenAI's Gemini-Pro to identify the most engaging parts of the video.
• Speaker Detection: Detects speakers in the video.
• Vertical Cropping: Crops the highlighted sections vertically, making them perfect for shorts.
• Caching System:
  • Stores processed video data in SQLite database
  • Caches transcriptions to avoid reprocessing
  • Saves highlight timestamps for quick retrieval
  • Improves processing speed for previously analyzed videos

Installation

Prerequisites

• Python 3.7 or higher
• FFmpeg
• OpenCV
• LangGraph
• SQLite3

Steps

1. Clone the repository:

   git clone https://github.com/SamurAIGPT/AI-Youtube-Shorts-Generator.git
   cd AI-Youtube-Shorts-Generator

2. Create a virtual environment:

   python3.10 -m venv venv

3. Activate the virtual environment:

   source venv/bin/activate # On Windows: venv\Scripts\activate

4. Install the python dependencies:

   pip install -r requirements.txt

Configuration

1. Set up the environment variables. Create a .env file in the project root directory and add your API key from Google AI Studio (it's free):

   GOOGLE_API_KEY=your_key_here

Usage

1. Ensure your .env file is correctly set up with your API key.
2. Run the main script:

   python main.py

3. Enter either:
   • A YouTube URL to process a new video
   • A local file path to process a video from your system

The tool will:

• Check if the video has been processed before
• Use cached data if available
• Only perform necessary processing steps for new videos
• Store results for future use

Database Structure

The caching system uses SQLite with three main tables:

• videos: Stores video metadata and file paths
• transcriptions: Stores video transcription data
• highlights: Stores extracted highlight segments
• If you face any issues or missing files with that try to remove the .db file

Known Issues

• Face detection and vertical cropping may to be fixed

Contributing

Contributions are welcome! Please fork the repository and submit a pull request.

License

This project is licensed under the MIT License.