Amazon Rekognition Content Moderation

For CoStar, it is imperative that images uploaded to our platform comply with the terms of our end user agreement and do not contain inappropriate content, so that we can ensure an inclusive, safe, and data-driven user community. Amazon Rekognition's Content Moderation API enabled us to easily build a solution to automatically analyze all uploaded images, allowing us to efficiently deliver high-value products to our customers. Amazon Rekognition offers a suite pre-trained computer vision APIs, which along with content moderation, text detection, and object detection, help us further improve our product offerings by making the images we receive more discoverable and our community more inclusive. Amazon Rekognition allows us to move quickly and add AI smarts to our systems with its pre-trained models, helping us stay focused on delivering unique solutions to the real estate sector.

Every decision we make is backed by data and technology, considering various metrics to continually add ‘wow factors’ that help retain customers. AWS promotes a user-first culture, with intuitive cloud-native services that help us launch things fast without any dependencies. The various AWS technology offerings help us develop our prototypes and make them live very quickly, even at a massive scale. This gives us a competitive edge in the market, where speed is essential.

As a large, global platform, unwanted content is extremely risky to the health of our community and can alienate photographers. We use Amazon Rekognition's content moderation feature to find and properly flag unwanted content, enabling a safe and welcoming experience for our community. At Flickr's huge scale, doing this without Amazon Rekognition is nearly impossible. Now, thanks to content moderation with Amazon Rekognition, our platform can automatically discover and highlight amazing photography that more closely matches our members' expectations, enabling our mission to inspire, connect, and share.

A large number of images are posted on WEAR from our users every day, and it was necessary to check every image to ensure that it complied with the service guidelines. We built a solution based on Amazon Rekognition Content Moderation API that automatically inspects to analyze the content that users post and store in Amazon S3. Amazon Rekognition has helped us cut down the manual content review process by up to 40% by automatically analyzing images. We were also able to reduce escalating reviews to supervisors that would have slowed down content operations when a person could not determine if an image was appropriate or not.