An Ongoing Research About Networked Home Appliances Using Computer Vision.

Connected objects will make the smart home and previously only experimental scenarios of ubiquitous computing an everyday reality. Sensors, computing power and data storage are getting ever more cheaper and widely available. All home appliances are currently getting equipped with sensors and network capabilities [1]. Camera is one of the most feasible sensors for capturing multivariate data. We believe that using a camera equipped networked home appliance and currently available technology it is possible to extract the following information:

Recognising user demographics such as the approximate age, gender, mood, location, time and length of interaction.
Storing personalised settings for each user (e.g. coffee machine).
Computer vision based auto-login and child protection.
Keeping track of interaction history, recognising and correlating behavioural patterns.
Computer vision based motion detection for home security and user notification.
Automatic product recognition.

[1] http://www.bbc.co.uk/news/technology-25616575

In collaboration with Jaak Kaevats.