Keith Roscarel

How Are Our Customers Using Facial Recognition?

Blog Post created by Keith Roscarel on Feb 19, 2017

The science of facial recognition has its roots in the 1960s and has steadily progressed to being accepted as a mainstream technology. The primary use case has been 'find the bad person'. We see this evolving quickly as use cases are realising solid business improvements such as increased security, automation and enhance customer insight.

Our Live Face Match (LFM) is the primary Hitachi technology that is used in facial recognition.

 

We divide video analytics technologies, such as facial recognition, into three main categories:

 

  • Security Analytics - LFM is deployed mostly to search for persons of interest whether they are on the 'black list' or 'white' list. This tends to be either standalone or directly integrated with security systems.
  • Operational Analytics - LFM data is integrated with 'line of business systems' (e.g. Time and Attendance) to monitor and optimise operations
  • Business Analytics - LFM data is utilised to provide trending, provide input to p@redictive models and support Key Performance Objects of an organisation (e.g. facility design, usage analysis, increase company-wide productivity)

 

LFM screen shot.jpg

Real-time Live Face Match Events

 

Use Cases

 

These are some of the ways facial recognition is being deployed by customers:

 

  • Find Person of Interest - identifying and alerting when an individual is located.
  • Exclude/Monitor Undesirable Individuals - enforce exclusion of individuals who have been banned from venues, store chains etc.
  • Self Exclusion - In cases where patrons wish to self-exclude from gambling or alcohol serving venues.
  • Customer Relationship and Loyalty - Create a VIP experience by identifying key customers as they arrive through your door.
  • Track Visitors and Workers on Building sites - Only allow access to individuals who have been correctly inducted onto a site.
  • Two Factor Authentication - Only allow access to individuals with their face recognised and a valid access card.
  • Site Activity and Usage - Understand how your worksite and/or customers are consuming services on your site.

 

two factor authentication with LFM.jpg

Two Factor Authentication Workflow

 

Solution Architecture

Integrated Facial Recognition.jpg

To include facial recognition into a security and/or operational environment, a 'face match' transaction is consumed either by specific security solutions (such as HVS) or via a data integration platform (such as Pentaho). The Web API is real-time so a transaction can be delivered to your data platform in milliseconds. This is necessary for solutions such as Two Factor Authentication (e.g. access card + face match) and fraud detection.

 

Summary

Facial recognition is being deployed to realise significant benefits in organisations worldwide. If you're considering digital transformation initiatives, it may be a valuable solution to increase security, provide automation and gain unique insight as to how customers/employees are utilising venues/campuses.

Outcomes