Facial Recognition is an optional add-on to identify known people in images uploaded into the resource section of IntelligenceBank.
IntelligenceBank uses industry leading Artificial Intelligence and Machine Learning technology to offer this solution.
Most common use cases are to identify executives, staff, students, brand ambassadors, celebrities and other known people in images.
As part of the setup process, a new database will be configured in your platform.
This database will hold all relevant details about each known person and allows users with sufficient permissions to add new or modify existing persons.
The database can be configured to store additional details about each person, such as:
- Full name
- Internal reference code
Adding a new person automatically creates a folder for this person within your resource area.
Users should upload as many clear, frontal shots of each person into each folder as practical. We recommend uploading between 3-6 images showing the person from different angles. Different lighting conditions, glasses, sunglasses, hair coverings/hats should be considered and will improve accuracy.
Uploaded photos can be linked to the relevant person using the database. The system will analyze each linked image, create a token (which can't be reverse engineered to re-create the face) and train the AI model for this person.
All images uploaded from that point forward will be checked for any known persons which will be automatically linked to the database record of each person. We recommend this is limited to a specific folder which is likely going to contain images of people.
The name's of all persons found (and any other details identified in the setup process) will be also be added to the keyword tags of the image.
Clicking the keyword tag will find all images using the same tag. Users can also open the related database records directly from the image which will show all images of this person.
The system can currently find up to 100 people within the same image.
The minimum detectable face size is 36x36 pixels in an image no larger than 1920x1080 pixels. Images with dimensions higher than 1920x1080 pixels will need a proportionally larger minimum face size.
Please contact your Customer Success Manager or email us at email@example.com to learn more and discuss your specific integration requirements.
Please sign in to leave a comment.