Document history for Amazon Rekognition
The following table describes important changes in each release of the Amazon Rekognition Developer Guide. For notification about updates to this documentation, you can subscribe to an RSS feed.
-
Latest documentation update: June 15th, 2023
| Change | Description | Date |
|---|---|---|
Amazon Rekognition now supports new moderation labels and improved accuracy for content moderation of images | Amazon Rekognition's Content Moderation feature has been enhanced for improved accuracy, detection of new labels, and the ability to identify animated and/or illustrated content. | February 1, 2024 |
Amazon Rekognition now supports bulk image analysis | Amazon Rekognition now supports processing a large collection of images asynchronously by using a manifest file with the StartMediaAnalysisJob operation. | October 23, 2023 |
Amazon Rekognition now supports custom content moderation with adapters | Amazon Rekognition now supports enhanced accuracy of the DetectModerationLabels API by using adapters that extend capabilities of existing Rekognition deep learning models. | October 12, 2023 |
Rekognition now supports user vectors with collections | Rekognition face collections now support the creation of user vectors. User vectors aggregate multiple face vectors of the same user, improving accuracy with more robust depictions of a user. | June 12, 2023 |
Actions for involving managing users have been added to the following managed policies: AmazonRekognitionReadOnlyAccess | Amazon Rekognition added the following actions to the
| June 12, 2023 |
Amazon Rekognition Image can now infer gaze direction | Improvements have been made to Amazon Rekognition Image's face detection operations, which can now infer gaze direction in a detected face. | May 31, 2023 |
Rekognition content moderation API improved | Rekognition improved the content moderation model for image and video moderation. The improvement significantly expands the detection of explicit, violence, and suggestive content. Customers can now detect explicit and violent content with higher accuracy to improve the end-user experience, protect their brand identity, and ensure that all content complies with their industry regulation and policies. | May 9, 2023 |
Amazon Rekognition Image can now detect occluded faces | Amazon Rekognition Image can now detected the occlusion of faces. A new FaceOccluded attribute is returned by Amazon Rekognition Image's DetectFaces and IndexFaces APIs, which indicates if the face in an image is partially captured or not fully visible due to overlapping objects, clothing, and body parts. | May 5, 2023 |
Rekognition can now detect face liveness | Amazon Rekognition Video can now be used to detect liveness in a video, verifying that a user in front of a camera is physically present. The Face Liveness detector also detects spoof attacks presented to a camera or trying to bypass a camera. | April 11, 2023 |
Update to Amazon Rekognition Video. | Amazon Rekognition Video can now detect more labels and return more information about attributes of images and labels. The GetLabelDetection API now returns information about aliases and categories. Returned label information can be filtered with inclusive and exclusive filter options. Results can be aggregated by Timestamps or video Segments. | December 7, 2022 |
Update to Amazon Rekognition Image. | Amazon Rekognition Image can now detect more labels and it now returns more information about attributes of images and labels. The DetectLabels API now returns information about aliases, categories, and image properties like dominant colors. Returned label information can be filtered with inclusive and exclusive filter options. | November 11, 2022 |
Actions for ProjectPolicy and Custom Labels Model Copy have been added to the following managed policies: AmazonRekognitionReadOnlyAccess | Amazon Rekognition added the following actions to the
| July 21, 2022 |
Actions for ProjectPolicy and Custom Labels Model Copy have been added to the following managed policies: AmazonRekognitionFullAccess, AmazonRekognitionCustomLabelsFullAccess | Rekognition added the following actions to the
| July 21, 2022 |
Amazon Rekognition Video can now detect labels in streaming video | Amazon Rekognition Video can detect labels such as pets and packages in streaming video. This
is done using | April 28, 2022 |
API reference moved out of the Amazon Rekognition developer guide | The Amazon Rekognition API reference is now available at Amazon Rekognition API Reference. | February 24, 2022 |
Dataset management update for the following managed policies: AWS managed policy: AmazonRekognitionReadOnlyAccess,AWS managed policy: AmazonRekognitionFullAccess,AWS managed policy: AmazonRekognitionCustomLabelsFullAccess | Amazon Rekognition added the following actions to the
| November 1, 2021 |
A new node in the table of contents shows Amazon Rekognition examples hosted on GitHub | Updated code examples from the AWS Code Examples Repository now show up in a separate node in the Amazon Rekognition developer guide for easier access. | October 22, 2021 |
Amazon Rekognition can detect black frames and primary program content in video segments | Amazon Rekognition can identify black frames, color bars, opening credits, end credits,
studio logos, and primary program content as technical cues in a video using the
| June 7, 2021 |
Dataset management update for the following managed policies: | You can use the Amazon Rekognition | May 21, 2021 |
Tagging update for AmazonRekognitionReadOnlyAccess and AmazonRekognitionFullAccess | Rekognition added new tagging actions to the
| April 2, 2021 |
Amazon Rekognition now supports tagging | You can now use tags to identify, organize, search for, and filter Amazon Rekognition collections, stream processors, and Custom Labels models. | March 25, 2021 |
Amazon Rekognition can now detect personal protect equipment | Amazon Rekognition can now detect hand covers, face covers, and head covers on persons in an image. | October 15, 2020 |
Amazon Rekognition has new content moderation categories | Amazon Rekognition content moderation categories now include 6 new categories: Drugs, Tobacco, Alcohol, Gambling, Rude Gestures, and Hate Symbols. | October 12, 2020 |
New tutorial for displaying Amazon Rekognition Video results from Kinesis Video Streams locally | You can display the output of Amazon Rekognition Video from a streaming video in Kinesis Video Streams in a local video feed. | July 20, 2020 |
New Amazon Rekognition tutorial for using Gstreamer | Using Gstreamer, you can ingest a livestream video from a device camera source to Amazon Rekognition Video through Kinesis Video Streams. | July 17, 2020 |
Amazon Rekognition now supports segmentation of stored videos | With the asynchronous Amazon Rekognition Video segmentation API you can detect black frames, color bars, end credits, and shots in stored videos. | June 22, 2020 |
Amazon Rekognition now supports Amazon VPC endpoint policies | By specifying a policy you can restrict access to an Amazon Rekognition Amazon VPC endpoint. | March 3, 2020 |
Amazon Rekognition now supports the detection of text in stored videos | You can use the Amazon Rekognition Video API to asynchronously detect text in a stored video. | February 17, 2020 |
Amazon Rekognition now supports Augmented AI (Preview) and Amazon Rekognition Custom Labels | With Amazon Rekognition Custom Labels you can detect specialized objects, scenes, and concepts in images by creating your own machine learning model. DetectModerationLabels now supports Amazon Augmented AI (Preview). | December 3, 2019 |
Amazon Rekognition now supports AWS PrivateLink | With AWS PrivateLink you can establish a private connection between your VPC and Amazon Rekognition. | September 12, 2019 |
Amazon Rekognition face filtering | Amazon Rekognition adds enhanced face filtering support to the IndexFaces API operation, and introduces face filtering for the CompareFaces and SearchFacesByImage API operations. | September 12, 2019 |
Amazon Rekognition Video examples updated | Amazon Rekognition Video example code updated to create and configure the Amazon SNS topic and Amazon SQS queue. | September 5, 2019 |
Ruby and Node.js examples added | Amazon Rekognition Image Ruby and Node.js examples added for synchronous label and face detection. | August 19, 2019 |
Unsafe content detection updated | Amazon Rekognition unsafe content detection can now detect violent content. | August 9, 2019 |
GetContentModeration operation updated | GetContentModeration now returns the version of the moderation detection model used to detect unsafe content. | February 13, 2019 |
GetLabelDetection and DetectModerationLabels operations updated | GetLabelDetection now returns bounding box information for common objects and a hierarchical taxonomy of detected labels. The version of the model used for label detection is now returned. DetectModerationLabels now returns the version of the model used for detecting unsafe content. | January 17, 2019 |
DetectFaces and IndexFaces operation updated | This release updates the DetectFaces and IndexFaces operation. When the Attributes input parameter is set to ALL, the face location landmarks includes 5 new landmarks: upperJawlineLeft, midJawlineLeft, chinBottom, midJawlineRight, upperJawlineRight. | November 19, 2018 |
DetectLabels operation updated | Bounding boxes are now returned for certain objects. A hierarchical taxonomy is now available for labels. You can now get the version of the detection model used for detection. | November 1, 2018 |
IndexFaces operation updated | With IndexFaces you can now use the QualityFilter input parameter to filter out faces detected with low quality. You can also use the MaxFaces input parameter to reduce the number of faces returned based on the quality of the face detection, and the size of the detected face. | September 18, 2018 |
DescribeCollection operation added | You can now get information about an existing collection by calling the DescribeCollection operation. | August 22, 2018 |
New Python examples | Python examples have been added to the Amazon Rekognition Video content along with some content reorganization. | June 26, 2018 |
Updated content layout | The Amazon Rekognition Image content has been reorganized along with new Python and C# examples. | May 29, 2018 |
Amazon Rekognition supports AWS CloudTrail | Amazon Rekognition is integrated with AWS CloudTrail, a service that provides a record of actions taken by a user, role, or an AWS service in Amazon Rekognition. For more information, see Logging Amazon Rekognition API Calls with AWS CloudTrail. | April 6, 2018 |
Analyze stored and streaming videos. New table of contents | For information about analyzing stored videos, see Working with Stored Videos. For information about analyzing streaming videos, see Working with Streaming Videos. The table of contents for the Amazon Rekognition documentation has been rearranged to accomodate image and video operations. | November 29, 2017 |
Text in image and face detection models | Amazon Rekognition can now detect text in images. For more information, see Detecting Text. Amazon Rekognition introduces versioning for the face detection deep learning model. For more information, see Model Versioning. | November 21, 2017 |
Celebrity recognition | Amazon Rekognition can now analyze images for celebrities. For more information, see Recognizing Celebrities. | June 8, 2017 |
Image moderation | Amazon Rekognition can now determine if an image contains explicit or suggestive adult content. For more information, see Detecting Unsafe Content. | April 19, 2017 |
Age range for detected faces. Aggregated Rekognition metrics pane | Amazon Rekognition now returns the estimated age range, in years, for faces detected by the Rekognition API. For more information, see AgeRange. The Rekognition console now has a metrics pane showing activity graphs for an aggregate of Amazon CloudWatch metrics for Rekognition over a specified period of time. For more information, see Exercise 4: See Aggregated Metrics (Console). | February 9, 2017 |
New service and guide | This is the initial release of the image analysis service, Amazon Rekognition, and the Amazon Rekognition Developer Guide. | November 30, 2016 |
