Amazon Rekognition Face Liveness launches accuracy improvements and new challenge setting for improved UX

Today, AWS announces accuracy improvements and new settings for Amazon Rekognition Face Liveness. Amazon Rekognition Face Liveness is a feature of Amazon Rekognition that detects in real-time whether real users, not bad actors using spoofs, can access your services.

Customers across financial, gig economy, telecommunications, healthcare, and social media use Rekognition Face Liveness detection for workflows such as onboarding, authentication, and bot detection. Until now, Rekognition Liveness only offered a single experience with the 'FaceMovementAndLightChallenge' setting, which delivers the highest accuracy by requiring users to move their face toward the screen and hold still for a series of flashing lights. With this launch, the new 'FaceMovementChallenge' setting reduces the check time by 3 seconds by eliminating the flashing lights. While 'FaceMovementAndLightChallenge' remains the best setting to maximize accuracy, 'FaceMovementChallenge' allows customers to prioritize faster liveness checks when appropriate. For additional flexibility, 'FaceMovementChallenge', allows users to complete checks using the front or back facing camera. Lastly, this update also delivers improved accuracy across both settings to aid with fraud mitigation.

The new Face Liveness settings are available in all AWS commercial regions where Rekognition Liveness is offered at no additional cost. Customers can enable the 'FaceMovementChallenge' setting in the CreateFaceLivenessSession API call.

To get started with the new settings, visit the Amazon Rekognition Face Liveness page or refer to the Amazon Rekognition Developer Guide.