UPDATED 23:19 EDT / APRIL 24 2019

AI

Amazon updates SageMaker Ground Truth to simplify dataset labeling

Amazon Web Services Inc. made machine learning training tasks a little easier Wednesday with an update to its Amazon SageMaker Ground Truth technology launched late last year.

The technology is a set of tools designed to be used with Amazon’s SageMaker service, which developers use to build machine learning models for predictive and analytical applications. The idea with the Ground Truth tools is to make it easier to label datasets that are used to train these machine learning models.

With Amazon SageMaker Ground Truth, developers can build more accurate datasets to train models for tasks including image classification, object detection, semantic segmentation and others. Essentially what the service does is provide workflows and interfaces that enable labeling tasks to be automated. And there’s good reason to want to do this, Amazon says, since automating labeling tasks can reduce costs by up to 70% in some cases.

Here’s how it works:

product-page-diagram_samurai_how-it-works-2-bc19de267c29570783c4add8bb2286ee584fcfbc

Amazon is introducing a host of new features to SageMaker Ground Truth, many of which were built due to demand from existing users, AWS Technical Evangelist Julien Simon wrote in a blog post.

The updates include a new “job chaining” feature that should help developers save time because it allows them to run subsequent machine learning labeling jobs, using the output of previous jobs.

“Basically, they want to chain together labeling jobs using the outputted labeled dataset (and outputted ML model if automated data labeling was enabled),” Simon explained. “For example, they may run an initial job where they identify if humans exist in an image, and then they may want to run a subsequent job where they get bounding boxes drawn around the humans.”

There’s also a new job tracking feature that allows developers to track the progress of labeling jobs in real time. Long-lived jobs, meanwhile, enables human experts to be used as labelers and update models on a periodic basis. Finally, there’s a new dynamic custom workflow feature, which allows additional context to be added to the source data, meaning developers can add input from previous labeling jobs.

In addition to the updates, Amazon said it’s now working with new service providers, including Vivetic and SmartOne, to add support for data labeling in French, German and Spanish. Amazon SageMaker Ground Truth is also being made available in more regions, including Virginia, Ohio, Ireland, Tokyo and Sydney.

Photo: Tony Webster/Flickr

A message from John Furrier, co-founder of SiliconANGLE:

Your vote of support is important to us and it helps us keep the content FREE.

One click below supports our mission to provide free, deep, and relevant content.  

Join our community on YouTube

Join the community that includes more than 15,000 #CubeAlumni experts, including Amazon.com CEO Andy Jassy, Dell Technologies founder and CEO Michael Dell, Intel CEO Pat Gelsinger, and many more luminaries and experts.

“TheCUBE is an important partner to the industry. You guys really are a part of our events and we really appreciate you coming and I know people appreciate the content you create as well” – Andy Jassy

THANK YOU