https://github.com/jaxony/turktool

Modern React app for bounding box annotation on mturk
https://github.com/jaxony/turktool

amazon-mechanical-turk bounding-boxes deep-learning javascript labeling object-detection react

Last synced: 4 days ago
JSON representation

Modern React app for bounding box annotation on mturk

• Host: GitHub
• URL: https://github.com/jaxony/turktool
• Owner: jaxony
• License: mit
• Created: 2017-11-12T10:23:16.000Z (about 7 years ago)
• Default Branch: master
• Last Pushed: 2019-12-06T19:24:01.000Z (almost 5 years ago)
• Last Synced: 2024-08-01T22:46:17.346Z (3 months ago)
• Topics: amazon-mechanical-turk, bounding-boxes, deep-learning, javascript, labeling, object-detection, react
• Language: JavaScript
• Homepage: https://jaxony.github.io/turktool
• Size: 895 KB
• Stars: 55
• Watchers: 3
• Forks: 17
• Open Issues: 2
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-dataset-tools - Turktool - Modern React app for scalable bounding box annotation of images (Labeling Tools / Images)

README

        
# turktool

`turktool` is a slick, modern React app for scalable bounding box annotation of images. You can connect it to Amazon Mechanical Turk and/or a custom backend to fulfill your deepest data labelling desires.

This project was bootstrapped with `create-react-app`. You can find the original (and incredibly useful) README.md [here](https://github.com/facebookincubator/create-react-app/blob/master/packages/react-scripts/template/README.md).

Check out the live [demo](https://jaxony.github.io/turktool/) hosted on GitHub Pages.

## Features

- Redux undo and redo of bounding boxes using keyboard

- Crosshairs for maximum accuracy

- Toggling crosshairs using keyboard

- POSTing annotations to Mechanical Turk via a form

- GETting image from a URL

## Non-features

- Object classification is not support (I treat this as a separate task)