Augmentor: An Image Augmentation Library for Machine Learning

TL;DR

Augmentor is a versatile image augmentation library in Python and Julia that enhances machine learning models by providing a flexible, pipeline-based approach for generating diverse augmented images, including advanced distortions.

Contribution

It introduces a high-level API for image augmentation with both standard and advanced features, facilitating improved data diversity for machine learning.

Findings

Supports standard and advanced augmentation techniques

Enables runtime sampling from augmented image distributions

Provides user-friendly API for effective data augmentation

Abstract

The generation of artificial data based on existing observations, known as data augmentation, is a technique used in machine learning to improve model accuracy, generalisation, and to control overfitting. Augmentor is a software package, available in both Python and Julia versions, that provides a high level API for the expansion of image data using a stochastic, pipeline-based approach which effectively allows for images to be sampled from a distribution of augmented images at runtime. Augmentor provides methods for most standard augmentation practices as well as several advanced features such as label-preserving, randomised elastic distortions, and provides many helper functions for typical augmentation tasks used in machine learning.