MuseBox

Face Analysis

The ML models for Face Analysis are in charge of extracting features from the face of a person which can be used to profile it. 

For this reason MuseBox is able to analyze the main characteristics of a person, such as its age, ornaments, expressions and face characteristics (i.e. how the eyes are).

Those analytics can be used in various applications where the analysis of the main features of the different people are very important. Possible applications are (in combination with the other components of MuseBox):

  • Surveillance
  • Smart Advertising (fit for every specific person)
  • Crowd analytics (i.e. during an event)
  • Attention Detection
  • Behaviour Analysis
  • Smart Retailing

Supported tasks

Face Detection
It detects the faces in a scene up to 32×32 pixel

Face Landmarking
Given a face bounding box, it extracts the 98 relevant points of a face 

Face Recognition
Given a face bounding box and a database of faces, it recognizes a face in a scene

Face Recognition with Mask
Given a face bounding box and a database of faces, it recognizes a face in a scene

Glasses Detection
Given a face bounding box, it detects whether a person wear glasses or not

Mask Detection
Given a face bounding box, it detects whether a person wear a mask or not

Gender Detection
Given a face bounding box, it determines if the person is female or male

Eyes and Blinking Detection
Given a face bounding box, it detects where are the eyes and it determines if they are blinking or not

Age Detection
Given a face bounding box, it determines the age of a person (do not use with girls or at least try to use our gender detector first to be sure 🙂 )

Emotion Detection
Given a face bounding box, it determines the emotion of a person (neutral, happy, sad, angry, surprise)