Searching for users (image)

SearchUsersByImagesearches the specified CollectionID for users in a collection that match the largest face detected in a supplied image. By default, SearchUsersByImage returns UserIDs for which the similarity score is greater than 80%. The similarity indicates how closely the UserID matches the largest face detected in the supplied image. If multiple UserIDs are returned, they are listed in order of highest similarity score to lowest. Optionally, you can use the UserMatchThreshold to specify a different value. For more information, see Managing users in a collection.

To search users by image (SDK)

1. If you haven't already:

   1. Create or update a user with AmazonRekognitionFullAccess permissions. For more information, see Step 1: Set up an AWS account and create a User.

   2. Install and configure the AWS CLI and the AWS SDKs. For more information, see Step 2: Set up the AWS CLI and AWS SDKs.

2. Use the following examples to call the SearchUsersByImage operation.

   Java

   This Java example searchers the users in a collection based off an inpute image, using the SearchUsersByImage operation.

   import com.amazonaws.services.rekognition.AmazonRekognition;
   import com.amazonaws.services.rekognition.AmazonRekognitionClientBuilder;
   import com.amazonaws.services.rekognition.model.Image;
   import com.amazonaws.services.rekognition.model.S3Object;
   import com.amazonaws.services.rekognition.model.SearchUsersByImageRequest;
   import com.amazonaws.services.rekognition.model.SearchUsersByImageResult;
   import com.amazonaws.services.rekognition.model.UserMatch;
   
   
   public class SearchUsersByImage {
       //Replace bucket, collectionId and photo with your values.
       public static final String collectionId = "MyCollection";
       public static final String s3Bucket = "bucket";
       public static final String s3PhotoFileKey = "input.jpg";
   
       public static void main(String[] args) throws Exception {
   
           AmazonRekognition rekognitionClient = AmazonRekognitionClientBuilder.defaultClient();
   
   
           // Get an image object from S3 bucket.
           Image image = new Image()
                   .withS3Object(new S3Object()
                           .withBucket(s3Bucket)
                           .withName(s3PhotoFileKey));
   
           // Search collection for users similar to the largest face in the image.
           SearchUsersByImageRequest request = new SearchUsersByImageRequest()
                   .withCollectionId(collectionId)
                   .withImage(image)
                   .withUserMatchThreshold(70F)
                   .withMaxUsers(2);
   
           SearchUsersByImageResult result =
                   rekognitionClient.searchUsersByImage(request);
   
           System.out.println("Printing search result with matched user and similarity score");
           for (UserMatch match: result.getUserMatches()) {
               System.out.println(match.getUser().getUserId() + " with similarity score " + match.getSimilarity());
           }
       }
   }

   AWS CLI

   This AWS CLI command searches the users in a collection based on an input image, with the SearchUsersByImage operation.

   aws rekognition search-users-by-image --image '{"S3Object":{"Bucket":"s3BucketName","Name":"file-name"}}' --collection-id MyCollectionId --region region-name

   Python

   The following example searches the users in a collection based on an input image, with the SearchUsersByImage operation.

   # Copyright 2018 Amazon.com, Inc. or its affiliates. All Rights Reserved.
   # PDX-License-Identifier: MIT-0 (For details, see https://github.com/awsdocs/amazon-rekognition-developer-guide/blob/master/LICENSE-SAMPLECODE.)
   
   import boto3
   from botocore.exceptions import ClientError
   import logging
   import os
   
   logger = logging.getLogger(__name__)
   session = boto3.Session(profile_name='profile-name')
   client = session.client('rekognition')
   
   def load_image(file_name):
       """
       helper function to load the image for indexFaces call from local disk
   
       :param image_file_name: The image file location that will be used by indexFaces call.
       :return: The Image in bytes
       """
       print(f'- loading image: {file_name}')
       with open(file_name, 'rb') as file:
           return {'Bytes': file.read()}
   
   def search_users_by_image(collection_id, image_file):
       """
       SearchUsersByImage operation with user ID provided as the search source
   
       :param collection_id: The ID of the collection where user and faces are stored.
       :param image_file: The image that contains the reference face to search for.
   
       :return: response of SearchUsersByImage API
       """
       logger.info(f'Searching for users using an image: {image_file}')
       try:
           response = client.search_users_by_image(
               CollectionId=collection_id,
               Image=load_image(image_file)
           )
           print(f'- found {len(response["UserMatches"])} matches')
           print([f'- {x["User"]["UserId"]} - {x["Similarity"]}%' for x in response["UserMatches"]])
       except ClientError:
           logger.exception(f'Failed to perform SearchUsersByImage with given image: {image_file}')
           raise
       else:
           print(response)
           return response
   
   def main():
       collection_id = "collection-id"
       IMAGE_SEARCH_SOURCE = os.getcwd() + '/image_path'
       search_users_by_image(collection_id, IMAGE_SEARCH_SOURCE)
   
   if __name__ == "__main__":
       main()

SearchUsersByImage operation request

The request for SearchUsersByImage includes the the collection to search in and the source image location. In this example, the source image is stored in an Amazon S3 bucket ( S3Object ). Also specified are the maximum number of users to return ( MaxUsers ) and the minimum confidence that must be matched for a user to be returned ( UserMatchThreshold).

{
    "CollectionId": "MyCollection",
    "Image": {
        "S3Object": {
            "Bucket": "bucket",
            "Name": "input.jpg"
        }
    },
    "MaxUsers": 2,
    "UserMatchThreshold": 99
}
            

SearchUsersByImage operation response

The response for SearchUsersByImage includes a FaceDetail object for the SearchedFace, as well as a list of UserMatches with the UserId, Similarity, and UserStatus for each. If the input image contained more than one face, a list of UnsearchedFaces will be also returned.

{
    "SearchedFace": {
        "FaceDetail": {
            "BoundingBox": {
                "Width": 0.23692893981933594, 
                "Top": 0.19235000014305115, 
                "Left": 0.39177176356315613, 
                "Height": 0.5437348484992981
            }
        }
    }, 
    "UserMatches": [
        {
            "User": {
                "UserId": "demoUser1", 
                "UserStatus": "ACTIVE"
            }, 
            "Similarity": 100.0
        }, 
        {
            "User": {
                "UserId": "demoUser2", 
                "UserStatus": "ACTIVE"
            }, 
            "Similarity": 99.97946166992188
        }
    ], 
    "FaceModelVersion": "6", 
    "UnsearchedFaces": [
        {
            "FaceDetails": {
                "BoundingBox": {
                    "Width": 0.031677018851041794, 
                    "Top": 0.5593535900115967, 
                    "Left": 0.6102562546730042, 
                    "Height": 0.0682177022099495
                }
            }, 
            "Reasons": [
                "FACE_NOT_LARGEST"
            ]
        }, 
        {
            "FaceDetails": {
                "BoundingBox": {
                    "Width": 0.03254449740052223, 
                    "Top": 0.6080358028411865, 
                    "Left": 0.516062319278717, 
                    "Height": 0.06347997486591339
                }
            }, 
            "Reasons": [
                "FACE_NOT_LARGEST"
            ]
        }
    ]
}