-
Book Overview & Buying
-
Table Of Contents
-
Feedback & Rating

Computer Vision on AWS
By :

Amazon Rekognition stores facial information inside server-side containers known as collections. Collections represent a logical grouping of Face Metadata, not the original image of the person. It supports operations for indexing, listing, searching, and deleting faces.
There’s no charge for collections; each holds 10 million Face Metadata, and you only pay for the aggregate Faces Metadata Storage ($0.00001/per facial metadata per month). For example, storing one million faces costs $1 per month, regardless of the total spanning collections.
You can create a collection using the AWS Command Line Interface (AWS CLI) v2 or the AWS SDK. When you invoke the CreateCollection
API, it only requires a name and returns instantly:
$ aws rekognition create-collection --collection-id \ "HelloWorld"
This command will report the following output:
{ «StatusCode": 200, «CollectionArn...