VidAI-Gen

A comprehensive tool for automatically downloading videos and adding interesting subtitles in Arabic, Persian, English, or any language to videos. It can also add a moving logo and create descriptions, titles, and thumbnails for videos—everything you need to work as a social creator! Now, all tasks are automatic, so feel free and enjoy it!

Features

• Automatic video transcription using Whisper models
• Translation between multiple languages (over 50 supported languages)
• YouTube video downloading with automatic processing
• Custom caption styling with font, color, and size options
• Word-by-word highlighting for better readability
• Special RTL (Right-to-Left) language support for Arabic and Persian
• Configurable shadow and stroke effects for subtitle visibility
• Pipeline-based processing for easy workflow automation
• Logo overlay capabilities with customizable positioning
• AI-powered title and description generation for content

Installation

Prerequisites

• Python 3.9+
• FFmpeg
• Internet connection for model downloads

Setup

1. Clone the repository:

git clone https://github.com/m15kh/VidAI-Gen.git
cd VidAI-Gen

2. Install the required dependencies:

pip install -r requirements.txt

3. Download the required models (will be done automatically on first run)

Usage

Basic Usage

Edit the configuration file at config/config.yaml with your video path or YouTube URL, then run:

python main.py

Using the Pipeline

The pipeline allows you to process videos through downloading, transcription, and captioning in one step:

python -m scripts.src.pipeline

Processing YouTube Videos

To download and process a YouTube video:

1. Set the video URL in config.yaml:

video_path: "https://youtube.com/shorts/kJ23zVr3HkE?si=M1O4CTvycVwbnZwR"
subtitle_path: null  # Let the system generate subtitles

2. Run the pipeline script

Processing Local Videos

To process a video you already have:

1. Set the video file path in config.yaml:

video_path: "/path/to/your/video.mp4"
subtitle_path: null  # Let the system generate subtitles

2. Run the pipeline script

Configuration

Configuration is done through the config/config.yaml file:

# Main configuration
video_path: "https://youtube.com/shorts/kJ23zVr3HkE?si=M1O4CTvycVwbnZwR"
subtitle_path: null  # Set to a JSON file path to use existing subtitles
output_dir: "output"
debug_mode: True

# YouTube download settings
youtube:
  quality: "best"  # Options: "best", "worst", or format code
  ratio: "9:16"    # Common values: "16:9" (landscape), "9:16" (portrait/shorts)

  output:
    filename: null  # Auto-generated if null
    merge: true     # Merge all segments into single video

# Subtitle processing settings
process_subtitle:
  enabled: true
  model: "large"    # Choices: tiny, base, small, medium, large
  verbose: false
  task: "translate" # Choices: transcribe, translate
  language: "en"    # Source language (auto for automatic detection) 
  translate_to: 'fa_IR'  # Target language, null for no translation

# Video editor settings
video_editor:
  logo:
    enabled: false  # Enable/disable logo overlay

Supported Languages

Auto-Subtitle supports over 50 languages including:

• English (en_XX)
• Persian/Farsi (fa_IR)
• Arabic (ar_AR)
• Chinese (zh_CN)
• German (de_DE)
• Spanish (es_XX)
• Russian (ru_RU)
• Korean (ko_KR)
• French (fr_XX)
• Japanese (ja_XX)
• Portuguese (pt_XX)
• Turkish (tr_TR)
• Hindi (hi_IN)
• Italian (it_IT)
• Dutch (nl_XX)
• and many more...

Output Files

The system generates several files during processing:

1. JSON subtitle file: Contains time-aligned words and sentences
2. WAV audio file: Temporary extracted audio for processing
3. Final video file: Video with the captions burned in
4. Content description file: Optional AI-generated video titles and descriptions

Customizing Captions

You can customize the appearance of captions by modifying parameters in the configuration or directly via function calls:

captacity.add_captions(
    video_file=video_path,
    output_dir=output_dir_path,
    subtitle_path=subtitle_path,
    font="/path/to/your/font.ttf",
    font_size=50,
    font_color="white",
    stroke_width=2,
    stroke_color="black",
    shadow_strength=1.0,
    shadow_blur=0.8,
    highlight_current_word=True,
    word_highlight_color="blue",
    line_count=2,
    padding=50,
    print_info=True,
)

Caption Parameters

Parameter	Description	Default
font	Path to TTF font file	System default
font_size	Size of subtitle text	50
font_color	Color of subtitle text	"white"
stroke_width	Width of text outline	2
stroke_color	Color of text outline	"black"
shadow_strength	Intensity of drop shadow (0.0-1.0)	1.0
shadow_blur	Blur radius for shadow	0.8
highlight_current_word	Whether to highlight the current word	True
word_highlight_color	Color for highlighted words	"blue"
line_count	Maximum number of subtitle lines	2
padding	Bottom padding for subtitles	50

Project Structure

auto-subtitle/
├── config/                 # Configuration files
│   └── config.yaml         # Main configuration
├── data/                   # Input video files
├── output/                 # Generated subtitle and video files
├── scripts/
│   ├── models/
│   │   ├── captacity/      # Caption styling and rendering
│   │   ├── subtitle/       # Subtitle generation
│   │   └── youtube/        # YouTube video download functionality
│   └── src/
│       └── pipeline.py     # Main processing pipeline
├── main.py                 # Entry point for direct captioning
├── README.md               # This file
└── TODO                    # Future plans and development notes

Arabic/Persian Text Support

The project includes special handling for Right-to-Left (RTL) languages:

• Arabic text reshaping
• Bidirectional text handling
• Special word order processing

Roadmap

Phase 1

• Support for short English videos with subtitles and logo addition

Phase 2

• Support for Persian videos from YouTube with subtitles

Phase 3

• Upgrade models from LLama2 to LLama3

Phase 4

• Automatic upload to Instagram

Phase 5

• Automatic upload to YouTube

Contributing

1. Fork the repository
2. Create a feature branch: git checkout -b feature-name
3. Commit your changes: git commit -m 'Add feature'
4. Push to the branch: git push origin feature-name
5. Submit a pull request

License

[Include license information here]