Detecting and analyzing faces

Amazon Rekognition provides you with APIs you can use to detect and analyze faces in images and videos. This section provides an overview of the non-storage operations for facial analysis. These operations include functionalities like detecting facial landmarks, analyzing emotions, and comparing faces.

Amazon Rekognition can identify facial landmarks (e.g., eye position), detect emotions (e.g., happiness or sadness), and other attributes (e.g., glasses presence, face occlusion). When a face is detected, the system analyzes facial attributes and returns a confidence score for each attribute.

This section contains examples for both image and video operations.

For more information about using the Rekognition’s image operations, see Working with images.

For more information about using the Rekognition’s video operations, see Working with stored video analysis operations.

Note that these operations are non-storage operations. You can use storage operations and Face collections to save facial metadata for faces detected in an image. Later you can search for stored faces in both images and videos. For example, this enables searching for a specific person in a video. For more information, see Searching faces in a collection.

For more information, see the Faces section of Amazon Rekognition FAQs.