Welcome to Jupyter-Innotater’s documentation!

Annotate data including image bounding boxes inline within your Jupyter notebook in python.

Innotater is a Jupyter notebook extension (widget) for machine learning practitioners. Use Python code to load your training and test data, then interactively annotate the data by drawing bounding boxes, selecting classes from a drop-down, or through other Javascript widgets.

Now works in JupyterLab (2.0+ recommended).

Read a full description in Overview or try out example Jupyter notebooks in Binder:

Easy local Installation using pip.

Documentation Contents:

• Overview
• Examples
  • Single Examples
    • Images and Bounding Boxes
    • Multi-Classification of Images
    • Binary-Classification with Bounding Boxes for Images
    • Multi-Classification of Text Strings
  • Multiple Examples
    • Multiple Bounding Boxes (with single multi-classification)
    • Multiple Bounding Boxes AND Multi-classification
    • Grouping
  • Jupyter Notebook Examples
  • Full Tutorial of an example project
• Usage Reference
  • Innotater
  • Innotation subclasses
    • ImageInnotation
    • BoundingBoxInnotation
    • MultiClassInnotation
    • BinaryClassInnotation
    • TextInnotation
    • ButtonInnotation
  • Repeats and Grouping
    • RepeatInnotation
    • GroupedInnotation
• Installation
  • Install from PyPi (recommended)
  • Development install
• Requirements
• Contact for Feedback
• License

Innotater source code can be found on GitHub here.