Image Clustering
Automatically group similar images using KMeans, HDBSCAN, or GMM algorithms with state-of-the-art feature extractors.
Now in Active Development
Organize, Clean &
Analyze Image Datasets
A comprehensive Python toolkit for dataset curation, duplicate detection, quality control, and exploratory data analysis with state-of-the-art ML models.
Powerful Features
Everything you need to organize and analyze your image datasets
Modern ML Models
Leverage DINOv2, CLIP, ViT, ResNet, EfficientNet, and ConvNeXt for powerful feature extraction.
Duplicate Detection
Find and remove duplicate or near-duplicate images from your datasets efficiently.
Visual Grids
Generate beautiful visual grids of clustered images for easy inspection and analysis.
Dimensionality Reduction
Visualize high-dimensional features with PCA, UMAP, or t-SNE for exploratory analysis.
Auto Organization
Automatically organize images into cluster folders and export results to JSON.
Installation
Get started in seconds with pip
Basic Installation
Terminal
pip install imageatlasFull Installation (with all models)
Terminal
pip install imageatlas[full]Note on CLIP
If you wish to use the CLIP model, install it manually:
pip install git+https://github.com/openai/CLIP.git
Quick Start
Cluster your images in just a few lines of code
example.py
from imageatlas import ImageClusterer
# Initialize clusterer with state-of-the-art features
clusterer = ImageClusterer(
model='dinov2', # DINOv2, CLIP, ViT, ResNet, etc.
clustering_method='kmeans',
n_clusters=10,
device='cuda' # or 'cpu'
)
# Run clustering on your images
results = clusterer.fit("./path/to/images")
# Save results to JSON
results.to_json("./output/clustering_results.json")
# Create visual grids for each cluster
results.create_grids(
image_dir="./path/to/images",
output_dir="./output/grids"
)
# Organize images into cluster folders
results.create_cluster_folders(
image_dir="./path/to/images",
output_dir="./output/clusters"
)That's it! Your images are now clustered, visualized, and organized. 🎉
Documentation
Explore available models and algorithms
Feature Extraction
State-of-the-art models
- dinov2 - DINOv2 (ViT)
- clip - OpenAI CLIP
- vit - Vision Transformer
- resnet - ResNet (18-152)
- efficientnet - EfficientNet
- convnext - ConvNeXt
Clustering
Grouping algorithms
- kmeans - K-Means
- hdbscan - HDBSCAN
- gmm - Gaussian Mixture
Parameters: n_clusters, min_cluster_size, min_samples, n_components
Dim. Reduction
Visualization methods
- pca - PCA
- umap - UMAP
- tsne - t-SNE
Parameters: n_components, n_neighbors, min_dist, perplexity
Citation
If you use ImageAtlas in your research, please cite it
BibTeX
@software{imageatlas,
author = {Ahmad Javed},
title = {ImageAtlas: A Toolkit for Organizing, Cleaning and Analyzing Image Datasets},
year = {2024},
url = {https://github.com/ahmadjaved97/ImageAtlas}
}