capabilities.work

Enroll and Search a Face

A 3-step recipe with Amazon Rekognition. We can help you get access.

Recipe · 3 steps

Enroll and Search a Face

with Amazon Rekognition

Create a collection, index an enrollment image, and search by a query image.

  1. Create the face collection that will store indexed face vectors.
  2. Detect faces in the enrollment image and add them to the collection.
  3. Search the collection for faces matching the supplied query image.
What you'll need to have ready
  • collectionId: ID for the face collection to create and use.
  • enrollBucket: S3 bucket holding the enrollment image.
  • enrollName: S3 object key of the enrollment image to index.
  • externalImageId: External identifier to assign to faces detected in the enrollment image.
  • queryBucket: S3 bucket holding the query image to search with.
  • queryName: S3 object key of the query image.
  • faceMatchThreshold: Minimum confidence for a search match to be returned.
For your developer

Arazzo workflow amazon-rekognition:enroll-and-search-face, 3 operations: CreateCollection → IndexFaces → SearchFacesByImage. Specs on APIs.io.

More recipes with Amazon Rekognition

Celebrity Scene ContextCustom Labels and ModerateDetect then Compare FacesFace Liveness SessionLabel and Moderate an ImageQuality Gated EnrollmentReuse or Create Collection then EnrollText and Moderation ScreenVerify a Face Against a CollectionVideo Label Detection Job