Skip to content

A comprehensive image analysis system using AI models to automatically tag, categorize, and critique photographs. This project includes standalone scripts, a web GUI, and a Lightroom plugin.

License

Notifications You must be signed in to change notification settings

gnartS/ai-photo-analyzer-rtx

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

17 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

AI Image Analyzer v4.0 - Streamlined Desktop & Lightroom Workflow

AI Image Analyzer

A powerful, local-first, AI-powered image analysis tool for professional and enthusiast photographers, now with a completely rebuilt, simplified, and reliable Adobe Lightroom Classic integration.

This major update abandons the previous complex web-based plugin in favor of a robust, practical, and familiar "send-to-app" workflow, inspired by industry-standard tools like Topaz Photo AI.


🎯 Key Features

  • Standalone Desktop GUI: A clean, modern interface for analyzing folders of images.
  • New Lightroom Plugin: A simple, reliable plugin to send images directly to the desktop app.
  • Two Processing Modes:
    • Archive Mode: Comprehensive analysis of all images for a fully searchable archive.
    • Curated Mode: Fast, quality-focused analysis of only the best images.
  • Local-First Processing: Your images are processed on your machine, ensuring privacy and security.
  • Powerful AI Models: Supports local models like BakLLaVA for speed and Gemini for quality.
  • RTX Acceleration: Optimized for NVIDIA RTX GPUs for blazing-fast performance.
  • XMP Sidecar Metadata: Non-destructive metadata that works with Lightroom, Bridge, and Capture One.

🚀 New Simplified Workflow

The new architecture is designed for simplicity and reliability:

  1. Select in Lightroom: Choose images in the Library module.
  2. Send to Desktop App: Use the Plug-in Extras menu to send images to the AI Image Analyzer desktop app.
  3. Analyze: Use the full power of the desktop app to analyze your images in either Archive or Curated mode.
  4. Refresh in Lightroom: Once finished, use the Refresh Metadata command in Lightroom to instantly see your new ratings, keywords, and descriptions.

This workflow is faster, more reliable, and more powerful than the previous web-based plugin.


🛠️ Installation

1. Install the Desktop App

  • Windows: Download the latest release from the Releases page and run the installer.
  • macOS/Linux: Clone the repository and run the app from source (see below).

2. Install the Lightroom Plugin

  1. Open Lightroom Classic.
  2. Go to File > Plug-in Manager.
  3. Click Add and navigate to the ai-image-analyzer-v2.lrplugin folder in this repository.
  4. The plugin is now installed and will appear in the Plug-in Extras menu.

💻 Running from Source

If you prefer to run the desktop app from source:

  1. Clone the repository:

    git clone https://github.com/donwrightdesigns/ai-photo-analyzer-rtx.git
    cd ai-photo-analyzer-rtx
  2. Install dependencies:

    pip install -r requirements.txt
  3. Run the main application:

    python main.py

🤖 Supported AI Models

  • BakLLaVA (Local, Recommended): Incredibly fast on NVIDIA RTX GPUs, ideal for large batches.
  • Gemini (Cloud): High-quality analysis using Google's API, great for single images or small batches.
  • Ollama (Local): Supports a wide variety of local models, good for privacy-focused users.

📈 Project Cleanup & Refactoring

This release represents a major cleanup and refactoring of the project. The following have been removed:

  • Old Lightroom Plugin: The entire ai-image-analyzer.lrplugin has been deleted and replaced with the much simpler ai-image-analyzer-v2.lrplugin.
  • Web Interface: The Flask-based web GUI has been removed to focus on a single, high-quality desktop application.
  • Obsolete Scripts: Numerous old and experimental scripts have been deleted to clean up the codebase.
  • Outdated Documentation: All documentation related to the old plugin and web interface has been removed.

This leaves a clean, focused, and maintainable codebase centered around the desktop application and its new, streamlined Lightroom integration.

About

A comprehensive image analysis system using AI models to automatically tag, categorize, and critique photographs. This project includes standalone scripts, a web GUI, and a Lightroom plugin.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Perl 67.5%
  • Python 30.7%
  • Lua 1.8%