AutoTagger

This project is a Flask-based web application that automatically generates tags for uploaded images using a published preexisting model.

Features

• Upload single or multiple images with a POST request
• Adjustable tag score threshold
• Response in JSON or HTML format
• REST API for integration with other services

Setup

Installation

1. Clone the repository, create a virtual environment, activate venv, install requirements. Setup env variables if needed:

   MODEL_PATH=models/model.onnx
   PORT=8000
   DEBUG=False
   

2. Download the model and tags:

   mkdir -p models
   wget -O models/model.onnx https://huggingface.co/SmilingWolf/wd-v1-4-swinv2-tagger-v2/resolve/main/model.onnx
   wget -O models/selected_tags.csv https://huggingface.co/SmilingWolf/wd-v1-4-swinv2-tagger-v2/raw/main/selected_tags.csv
   

Running and Usage

1. Make sure your virtual environment is active.

2. Start the Flask server:

   python app.py
   

3. Access the application at http://localhost:8000 in your web browser to use the web interface for uploading and tagging images.

4. For API usage, send POST requests to /evaluate. Example using curl:

   curl -X POST -F "file=@/path/to/your/image.jpg" -F "threshold=0.5" -F "format=json" http://localhost:8000/evaluate
   

   This command uploads an image file, sets the tag threshold to 0.5, and requests JSON output.

   Parameters:

   • file: The image file to upload (can be specified multiple times for batch processing)
   • threshold: The minimum confidence score for tags (default: 0.1)
   • format: Output format, either 'json' or 'html' (default: 'json')

   The API will return a JSON array of objects, each containing the filename and its associated tags.