Trending: The most powerful console yet, or a refrigerator? Reactions to Microsoft’s new Xbox Series X
AWS CEO Andy Jassy delivers the keynote at the 2018 reInvent conference. (Amazon Web Services Photo)

Amazon is rolling out new features that make it easier for customers to search for specific items in large databases of images and videos. The tech giant said Monday it will make “Custom Labels” available next month, allowing Rekognition users without any machine learning experience to deploy the software more nimbly.

Custom Labels allow customers to train Rekognition to detect specific objects and scenes related to their use case. For example, a retailer who wants to identify all of the black boots in her online inventory could train the software with images of shoes that meet that description, so that it can detect them in a larger library of images. That type of custom sorting previously required machine learning expertise to build a model from scratch, an Amazon spokesperson said.

Amazon offered another use case for Custom Labels in a blog post published Monday. The company said an auto shop could use the tool to identify specific machine parts, like a turbocharger, from a library of images.

Customers can upload and label images like these to train Rekognition to find specific items. (Amazon Image)

The Rekognition updates come amid a heated and ongoing debate over the facial recognition applications of Amazon’s Rekognition software. The company has faced criticism from civil rights groups for selling Rekognition to law enforcement agencies.

The ACLU and MIT conducted studies that showed facial recognition software can amplify human biases by misidentifying women and people of color more frequently than their white male counterparts. Amazon disputes those studies, claiming they were not conducted with the confidence threshold that the company recommends for law enforcement applications of the software.

The ACLU continued its steady stream of criticism Monday. BuzzFeed News ran an editorial by ACLU analyst Jay Stanley defending the rights of protesters to wear masks to defend against “the increasing prevalence of new surveillance technologies, especially surveillance cameras and face recognition.”

In October, the ACLU filed a lawsuit against the Justice Department and FBI seeking documents related to their use of surveillance technology. Amazon and Microsoft, which also develops facial recognition software, are named in the lawsuit.

It’s not just the ACLU. The digital rights advocacy group Fight for the Future has partnered with musicians and other organizations to urge artists to pull their music from Amazon’s streaming platforms. Some artists have already done so to pressure the company to sever ties with U.S. Immigrations and Customs enforcement.

Amazon has sought to dispel concerns over its facial recognition software in blog posts and public comments.

“As in all probabilistic systems, the mere existence of false positives doesn’t mean facial recognition is flawed,” Amazon said in a blog post titled “The Facts on Facial Recognition with Artificial Intelligence.” “Rather, it emphasizes the need to follow best practices, such as setting a reasonable similarity threshold that correlates with the given use case.”

An Amazon spokesperson said Custom Labels could be used to analyze images or videos that include people but the feature isn’t designed to detect faces. Custom Labels is part of a separate API from Amazon Rekognition’s facial recognition capability.

Custom Labels will be available starting Dec. 3. Amazon says the NFL, and a marketing platform called VidMob, are already piloting the new features.

Like what you're reading? Subscribe to GeekWire's free newsletters to catch every headline

Comments

Job Listings on GeekWork

Find more jobs on GeekWork. Employers, post a job here.