Face recognition for customer service

Face recognition for Over the last ten years, face recognition systems have become an important fact of life in the field of biometric analysis. These systems are comparable with digital fingerprint and iris recognition systems, which people use today to make online payments or gain access to their mobile devices.

Blackstone Face technology identifies the distinctive or photometric features of the human face, using an algorithm that examines the size and shape of the eyes, the proportions of the nose and the thickness of the cheek and jawbones, and uses this information to recognize individual people in a video or photograph. An emerging trend in this software family is 3D facial recognition, which provides an even more accurate analysis and reduces the margin of error to a minimum.

Blackstone eIT Object Recognition Server – Facial Recognition

  • Identify a person from a digital image as input by searching against a huge database
  • Handling the most complex and large-scale scenarios, live and offline matching
  • Not foiled by face changes & wearing (age, head hair, helmets, head covers etc..)
  • Robust to face parts occlusions, illumination and viewing variations.
  • Using latest CNN networks in academia and production
  • Trained using two of the largest datasets (Facescrub and Casia-WebFace)
  • Analyzing and adapting our own developed algorithms to the specific domain demographic factors
  • Match images with very fast speed (< 2s for millions of images)
  • Tailored for customer needs and imaging environment (pre-processing methods)
  • Compares large dataset of images

Blackstone Artificial Neutral Networks (ANN)

  • A machine learning inspired by the human nervous system
  • Interconnected processing units called neurons
  • Multiple layers made up of neurons using the same activation function
  • A common type of activation functions is the Sigmoid function
  • Feedforward Neural Network are neurons with loops or cycles connecting them
  • Convolutional Neural Network (CNN) is a category of feedforward artificial neural networks
  • Applied in Image and Facial Recognition
  • Recurrent Neural Network (RNN) are ANNs with directed cycles in the connection

Images challenges in surveillance cameras

  • Illumination variations
  • Lighting, camera characteristics and control, skin reflectance properties
  • Viewing variation
  • Face parts occlusion
  • Facial expression/facial style
  • Low resolution images
  • Noisy Images
  • Images low quality when provided by night vision cameras

The integration of the digital world with Contact Center services is hugely beneficial to the activities of Transcom and its customers. It makes it possible, for example, to drastically reduce the time required for initial customer authentication and the length of calls in general (10% less according to Deloitte). In addition, during calls, face recognition software can interpret the customer’s facial expressions, increasing the quality and value of feedback and providing support for the customer’s Net Promoter Score assessment. Finally, video chat systems with face recognition are also an excellent sales tool. According to reports from Citrix, the increase in sales is about 34%..

Why Blackstone eIT Facial Recognition?

  • Own developed algorithms CNN based (State of the art deep learning method)
  • Work on low resolution Images
  • Facial expressions don’t foil recognition
  • Customizable solution : pre-processing methods are customizable depending on the customer imaging environment (high/medium/low resolution)
  • While other AI solutions are not production ready, our FR server is a Production Server in terms of deployment, scalability, reliability, and monitors
  • High accuracy on huge datasets
  • While face picture can foil other solutions recognition, our FR server can determine whether a face detected in video stream belongs to a real human or just a picture

HR Attendance Tracking

Event Checkin/Registration System.

Airline /Airport Auto Check-In

Access Control System

Robotics Facial Recognition

Robotics Emotional Scanner

Watch these two quick demo videos that showcase the power of chatbots, and explain how you can get started with your own process optimizing chatbot.

Customers of all sizes succeed with Blackstone eIT

Our Offices


Seattle WA

T:+1 (425) 283-3463