Classto

Classto is a Python library for building lightweight, browser-based tools to manually classify images into custom categories - ideal for preparing datasets or sorting visual content.

With just a few lines of Python, Classto spins up a local web interface built on Flask and styled with Tailwind CSS to let you quickly review, label, and organize images - right from your browser.

Interface Previews

   

Classto in Light and Dark Mode

 

Features

• Classify images with one click
• Supports custom category lists
• Images are moved into subfolders per label
• Optionally add unique filename suffixes
• CSV logging for ML dataset tracking
• Delete unwanted images during classification
• Dark mode toggle built in

Installation

You can install Classto via pip:

pip install classto

Quickstart

import classto as ct

app = ct.ImageLabeler(
    classes=["Cat", "Dog"],
    image_folder="images",          # Path to your images
    delete_button=True,             # Optional delete button
    suffix=True,                    # Add unique suffix to avoid conflicts
    log_to_csv=True                 # Save results to labels.csv
)

app.launch()

Then open your browser at http://127.0.0.1:5000.

Example Folder Structure

Place your images in a folder (e.g. images/) relative to your script:

project/
├── images/
│   ├── cat1.jpg
│   ├── cat2.jpg
│   ├── dog1.jpg
│   └── dog2.jpg
├── app.py

After classification, images are moved to:

project/
├── classified/
│   ├── Cat/
│   │   ├── cat1__K8dLs.jpg
│   │   └── cat2__a7JkL.jpg
│   ├── Dog/
│   │   ├── dog1__Xy4Tz.jpg
│   │   └── dog2__Zx9Pm.jpg
│   └── labels.csv

Parameters

• classes (List[str]): A list of categories for classification (e.g. ["Dog", "Cat"]).
• image_folder (str): Path to the folder containing the images to classify. Defaults to "images".
• delete_button (bool): If True, shows a delete button to remove images. Defaults to False.
• suffix (bool): If True, appends a random suffix to filenames to avoid overwriting. Defaults to False.
• log_to_csv (bool): If True, logs each classification into a CSV file. Defaults to False.
• shuffle (bool): If True, images are shown in random order. Defaults to False.

How it works

1. Classto loads images from your specified folder (e.g. images/)
2. Each classification moves the image to classified/<category>/
3. Optionally appends a suffix to the filename
4. If enabled, logs the action to a labels.csv file in classified/

CSV Logging Format

If log_to_csv=True is enabled, each classification is logged to a labels.csv file inside the classified/ directory. The file contains the following columns:

original_filename	new_filename	label	timestamp
img01.jpg	img01__4Fg7T.jpg	Cat	2025-08-19T10:58:00+00:00

• original_filename: The name of the image before classification.
• new_filename: The new name after suffixing (if enabled).
• label: The category selected during classification.
• timestamp: When the classification occurred, in ISO 8601 format (UTC).