Data Collection

Skeletal Recognition

TeleMetriX combines information about a target's emotional state with information about body position to create a detailed profile of a person's interaction with their environment.

Product\Brand Recognition

A custom implementation of Tensor Flow allows TeleMetriX to detect specific products and brands using the latest in machine learning.

Machine Learning & Artificial Intelligence

TeleMetriX is powered by Affectiva, the worlds most comprehensive database of Human Emotional information with over 6 million faces.

Affectiva combined demographic information (Gender, Age and Ethnicity) with detailed facial metrics containing emotional states to summarize the complex emotions exhibited human beings.

•Demographics (Age, Gender & Ethnicity)

  • 5 Ethnic Groups

  • 7 Age Groups

•Emotional States

  • 7 Emotional States

  • 20 Expressions

  • 13 Composite Emojis

These metrics are combined to create two unique composite emotions that assist the system with making rapid decisions about a subject state of mind.

  • Engagement - Interest Level

  • Valence - Positive or Negative Response.

A Learning System

With all of this technology, TeleMetriX is able to learn new information.  Emotions are currently detected with more than 85% accuracy.  Demographic information at more than 65%.  Product information is improving with each interaction with the system.

And because of this flexible open ended framework, TeleMetriX will improve over time, allowing you to add new levels of detection, based on your internal needs.