Accurate Context Sharing

Accurate Context Sharing is a demo I have made for the Bluewave project. It shows a way to accurately sharing the context using facial information in the context. In this demo, I have made a intelligence sign which knows who is the person in front of itself.

For this system to work, I require user to input its name and take a few photo of him or her in a app installed on the user’s phone. The phone then automatically share its context within a certain range by bluewave. Although the context is not accurately shared, but beacuse that the facial data is also shared, the service provider (intelligence sign in this scenerio) can easily build a facial recognition model consist of only nearby people and their other context. When the sign detected that a person is around, it can find the person’s name from the model. In this demo we share the user’s facial images, but it can easily be upgraded to only share a trained vector of user’s facial characterics instead to eliminate the privacy issue.

By utilizing this method, one can easily build service which require information of a specific user, while protect user’s privacy and require only little computional data for the service provider.