https://github.com/svetjeonline/image-tagger
This repository contains a Python script for a desktop application that generates English captions and keywords for images using AI (Hugging Face Transformers) and writes them to IPTC/XMP metadata.
https://github.com/svetjeonline/image-tagger
ai captions gui image image-processing metadata photoshop python stock tagger
Last synced: 4 months ago
JSON representation
This repository contains a Python script for a desktop application that generates English captions and keywords for images using AI (Hugging Face Transformers) and writes them to IPTC/XMP metadata.
- Host: GitHub
- URL: https://github.com/svetjeonline/image-tagger
- Owner: svetjeonline
- Created: 2025-04-24T12:05:18.000Z (5 months ago)
- Default Branch: main
- Last Pushed: 2025-04-24T12:27:11.000Z (5 months ago)
- Last Synced: 2025-05-19T00:17:38.506Z (5 months ago)
- Topics: ai, captions, gui, image, image-processing, metadata, photoshop, python, stock, tagger
- Language: Python
- Homepage:
- Size: 86.9 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# 📸 AI Image Tagger & Describer 🏷️
### *Smart metadata for your image library. Automated. Local. Fast.*
---
## 🚀 Overview
**AI Image Tagger & Describer** is a cross-platform desktop application that automatically generates **captions** and **keywords** for your images using AI — and embeds this data into **IPTC/XMP metadata**.
🔍 Ideal for photographers, archivists, and content creators who manage large image libraries.
🧠 Works offline — your images stay on your machine.
---
## 🌟 Key Features
- 🧠 **AI-Powered Captions & Keywords**
Uses Salesforce's BLIP model (Base or Large) via Hugging Face Transformers
- ✍️ **Embedded Metadata**
Writes data directly into IPTC and XMP fields inside image files
- 💻 **Cross-Platform**
Works on Windows, macOS, and Linux
- 🖼️ **User-Friendly Interface**
Features image list, preview panel, and editable metadata fields
- ⚙️ **Fully Customizable Settings**
Select AI model, device (CPU/GPU), stop words, and metadata options
- ⚡ **Batch Processing**
Tag and save metadata for multiple images in one go
- 📝 **Logging & Error Handling**
Includes detailed logs and friendly error messages
- 📦 **Virtual Environment Support**
Keeps your dependencies clean and isolated
---
## 🧰 Technologies Used
- `Python 3.9+`
- `CustomTkinter` for GUI
- `Hugging Face Transformers` (BLIP model)
- `PyTorch` (optional for GPU acceleration)
- `iptcinfo3` for IPTC metadata
- `python-xmp-toolkit` and `libexempi` for XMP metadata
---
## 🖥️ Application Interface
### 📂 Image List Panel
- Displays image thumbnails and file names
- Shows processing status indicators
- Checkboxes for batch selection
---
### 🖼️ Detail Panel
- Large preview of the selected image
- Editable caption and keywords
- Save button to update metadata
---
## ⚙️ How It Works
1. **📂 Load Folder** – The app scans a selected directory for supported image files (JPG, JPEG, TIFF).
2. **📖 Read Metadata** – IPTC and XMP metadata is extracted if available.
3. **🤖 AI Tagging** – BLIP generates English captions and keywords.
4. **✍️ Write Metadata** – Results are written back into the image files.
---
## 🔧 Setup Instructions
### ✅ Prerequisites
- **Python 3.9+**
- `pip` (included with Python)
- **Visual C++ Build Tools** (needed if `python-xmp-toolkit` install fails)
- **Exempi** (required for XMP support, auto-installed via Conda or manual)
---
### 🛠️ Installation Guide
```bash
# 1. Clone the repository
git clone https://github.com/yourusername/yourrepository.git
cd yourrepository
# 2. Create a virtual environment
python -m venv .venv
# 3. Activate the virtual environment
# Windows:
.venv\Scripts\activate
# macOS/Linux:
source .venv/bin/activate
# 4. Install dependencies
pip install -r requirements.txt
# 5. Run the application
python main.py