For machine learning models, we all know more data is better. For convolutional neural networks, image augmentation provides a straightforward way to expand your training dataset, by applying simple transformations to the images you already have. In this talk I’ll start by demoing the basic standard transforms using the albumentations python package, and work up to some more advanced strategies like CutMix and mixup. I will also discuss some findings of the RxRx1 kaggle competition that Recursion ran last summer, and how this demonstrated the power of these techniques when applied to our cellular image data.
Author Name: Pluralsight LIVE
Author Description:
Pluralsight LIVE is the ultimate gathering of industry experts, business leaders and change-makers. As Pluralsight’s annual user conference, LIVE is where technologists from around the world come together to look into the future and prepare for the challenges and opportunities ahead. Subject matter experts share their insights on emerging trends, innovative business leaders talk about their breakthrough strategies, and Pluralsight specialists show users how to leverage the platform to reach thei… more

Table of Contents

• Image Augmentation: A Practical Guide to Prevent Overfitting in Computer Vision
  28mins