Facial recognition is a built-in, opt-in feature in Acquia DAM, which automatically detects and groups faces in images. This feature enables asset administrators to efficiently manage people tagging, improve content findability, and maintain compliance with data privacy protection laws.
Prerequisites
Obtain access to the Admin app.
Obtain the Administrate Face Tags permission. This permission is automatically granted to Global admin roles.
For non-Global admin roles, users must also have the "Edit Assets' Metadata, Type, and Preview" permission.
Enable the facial recognition feature in the Admin app. This feature is included out-of-the-box and does not incur any additional cost.
Enabling facial recognition
Access the Admin app.
Navigate to Feature settings.
Enable the facial recognition feature.
Verify the permissions for relevant users.
Asset processing after feature activation
New uploads: Facial recognition automatically runs post-upload for all newly uploaded images.
Processing time: The system might take a few minutes to display the results.
Existing assets (Bulk Backfill) For assets that existed in your DAM before you enabled the facial recognition feature, administrator accounts can run a self-service bulk backfill process.
Best practice before processing:
Before you run a backfill, determine if any Upload Profiles must be exempt from facial recognition processing.
Navigate to Admin > Upload Settings > Upload profiles.
Select the desired profile.
Deselect Run facial recognition (images only) in the Post upload processing section.
Any assets originally uploaded through an exempt profile are not processed during the backfill.
Run the backfill:
1. Navigate to Admin > Features.
2. Under the enabled Facial Recognition feature, select Run Facial Recognition.
Processing status: A success dialog box opens to indicate that the system started the process for existing assets. Because processing takes time depending on the size of your library, the administrator account that enabled the feature receives an email notification after the process completes. After the process completes, the system tags identified faces in the Asset Digest.
How facial recognition works
Face detection
Uses Amazon Rekognition technology
Detects up to 100 faces per image
Requires minimum face size of 50x50 pixels
Creates bounding boxes around detected faces
Face comparison
Compares detected faces against existing processed faces
Groups similar faces together
Enables display name assignment
Maintains consistency across multiple assets
Processing states
New uploads are automatically processed.
Processing might take a few minutes.
Status indicators are displayed when facial recognition is running.
Results are displayed in the Facial recognition section of Asset Digest.
Managing people tags
Name new faces
Locate unnamed faces in Asset Digest.
Click Name this person.
Do one of the following:
Select an existing name from the dropdown.
Create a new name.
4. Verify using primary images.
5. Confirm the selection.
Correcting AI mistakes
Fixing unmatched faces
Click the person in Asset Digest.
Select from known people.
Use primary images for verification.
Create new names if needed.
Fixing incorrect grouping
Click the person in Asset Digest.
Select This is not [Name] option.
Confirm untag for single image.
Click Name this person.
Assign the correct name.
Understanding primary images
What is a primary image?
Each known person has an automatic primary image.
Primary images are set automatically.
Primary images are used for consistent identification across the platform.
Primary images help to verify the correct person during tagging.
How primary images work
Primary images are displayed on person buttons
Different from asset face (current image face detection)
Used for verification during tagging
Consistent across all instances of the person
Using primary images for verification
Check primary image when selecting known people
Verify identity matches before tagging
Create new names for different people with same name
For named people: FacialRecognition:{Name} or af:{Name}
For unnamed people: FacialRecognitionUnnamedPerson:true or afup:true
5. Advanced Search Page:
Dedicated fields for named people and unnamed people
Accessible through the advanced search UI
Provides visual interface for facial recognition search criteria
Editing a display name
Users with appropriate permissions can edit person names across the system. The edit name feature allows you to:
Update display names across all tagged images
Immediate updates in Asset Digest
Brief delay for search index updates
How does facial recognition work in Acquia DAM?
Overview
Facial recognition is a built-in, opt-in feature in Acquia DAM, which automatically detects and groups faces in images. This feature enables asset administrators to efficiently manage people tagging, improve content findability, and maintain compliance with data privacy protection laws.
Prerequisites
Obtain access to the Admin app.
Obtain the Administrate Face Tags permission. This permission is automatically granted to Global admin roles.
For non-Global admin roles, users must also have the "Edit Assets' Metadata, Type, and Preview" permission.
Enable the facial recognition feature in the Admin app. This feature is included out-of-the-box and does not incur any additional cost.
Enabling facial recognition
Access the Admin app.
Navigate to Feature settings.
Enable the facial recognition feature.
Verify the permissions for relevant users.
Asset processing after feature activation
New uploads: Facial recognition automatically runs post-upload for all newly uploaded images.
Processing time: The system might take a few minutes to display the results.
Existing assets (Bulk Backfill) For assets that existed in your DAM before you enabled the facial recognition feature, administrator accounts can run a self-service bulk backfill process.
Best practice before processing:
Before you run a backfill, determine if any Upload Profiles must be exempt from facial recognition processing.
Navigate to Admin > Upload Settings > Upload profiles.
Select the desired profile.
Deselect Run facial recognition (images only) in the Post upload processing section.
Any assets originally uploaded through an exempt profile are not processed during the backfill.
Run the backfill:
1. Navigate to Admin > Features.
2. Under the enabled Facial Recognition feature, select Run Facial Recognition.
Processing status: A success dialog box opens to indicate that the system started the process for existing assets. Because processing takes time depending on the size of your library, the administrator account that enabled the feature receives an email notification after the process completes. After the process completes, the system tags identified faces in the Asset Digest.
How facial recognition works
Face detection
Uses Amazon Rekognition technology
Detects up to 100 faces per image
Requires minimum face size of 50x50 pixels
Creates bounding boxes around detected faces
Face comparison
Compares detected faces against existing processed faces
Groups similar faces together
Enables display name assignment
Maintains consistency across multiple assets
Processing states
New uploads are automatically processed.
Processing might take a few minutes.
Status indicators are displayed when facial recognition is running.
Results are displayed in the Facial recognition section of Asset Digest.
Managing people tags
Name new faces
Locate unnamed faces in Asset Digest.
Click Name this person.
Do one of the following:
Select an existing name from the dropdown.
Create a new name.
4. Verify using primary images.
5. Confirm the selection.
Correcting AI mistakes
Fixing unmatched faces
Click the person in Asset Digest.
Select from known people.
Use primary images for verification.
Create new names if needed.
Fixing incorrect grouping
Click the person in Asset Digest.
Select This is not [Name] option.
Confirm untag for single image.
Click Name this person.
Assign the correct name.
Understanding primary images
What is a primary image?
Each known person has an automatic primary image.
Primary images are set automatically.
Primary images are used for consistent identification across the platform.
Primary images help to verify the correct person during tagging.
How primary images work
Primary images are displayed on person buttons
Different from asset face (current image face detection)
Used for verification during tagging
Consistent across all instances of the person
Using primary images for verification
Check primary image when selecting known people
Verify identity matches before tagging
Create new names for different people with same name
Acquia DAM