ML Unboxed: How to get started with labeling automation || WBN

Join us for another edition of ML Unboxed! You’ll learn about essential ML techniques and emerging concepts and walk away with the resources you need to put them into practice at your organization.

In this episode of ML Unboxed you will learn how to dramatically reduce labeling time and associated cost with model-assisted labeling. This hands-on demonstration will teach you how to use your own model, or an open source model of your choice, to help automate labeling.

Teams using model-assisted labeling have seen a savings of 50-70% in terms of annotation costs and can dramatically improve overall project efficiency. In this tutorial, we’ll walk through workflows to help you increase labeling velocity whether you’re actively iterating on production models or are just getting started on your AI journey. Example workflows include learning how to:

Use an open source model to accelerate your first batches of training data.
Use your own model to pre-label data so labeling teams can focus on adjusting annotations vs. starting from scratch.
Use the model of your choice to apply annotations for use in weak supervision workflows.

This end-to-end tutorial will walk you through best practices as well as hands-on notebooks you can use to automate labeling your own data.

Watch now

Just some of companies that Labelbox is working with to build AI applications:

Jenna Wang

Senior Product Manager, Labelbox

Featured Speakers

How to get started with labeling automation

Just some of companies that Labelbox is working with to build AI applications:

ML Unboxed: How to diagnose and improve model performance

January 12: 11am PT

January 13: 8am PT/11am ET/4pm GMT

ML UNBOXED

Mark Ghannam

Solutions Engineer, EMEA, Labelbox