Human facial expressions are consistent for everyone regardless of who they are and where they come from. Our face reveals so much about how we feel, our physical health, and the state of our lives. The smile is a universal example of happiness recognized by anyone on the planet regardless of language spoken or education level.

If you are happy, you smile, if you are sad, you frown, and you may or may not be conscious of your facial expressions. What if we all made the same expression when we are sick and what if that was specific to the type of ailment? Computer vision training for machine learning models to identify patterns in facial expression that are invisible to the naked eye is why we need data sets of people before confirmed infection, during (as well as the added benefits of social media grouping clusters of people who were together at the time of contagious but unknown infection.

We are all human and we share consistent patterns of facial expression. If we all smile when we are happy maybe we all make the same expression when we are infected with COVID-19 and we can’t see it with human eyes. Computer vision and ML training with algorithm catalogues and data-sets organized here we can maybe identify patterns using machine learning models that the human eye can’t recognize.

The number of infected people worldwide grows exponentially every day. This makes data-sets available across all nations from all sorts of people from all walks of life. If we can get those who have smartphones with auto backups (e.g. consent to share app data from WhatsApp, iCloud, google photos, Samsung cloud/, one drive, etc.) To give us access to their photos taken before and after they were sick we may be able to put together databases that we can use to train ML model(s) to identify the probability of infection based on a selfie and their maps time/date stamped maps and image metadata.