What You See is Where You Are - Localize a Mobile Device with its Camera and WiFi

Localizing users and devices in an in-door environment remains an unsolved practical problem, particularly if sub-meter accuracy and ubiquitous hardware are constraints of the use case to support. Although in recent years, there have been several efforts to provide localization services for indoor environments, however many of them are either not precise enough or require non-standard or non-wearable equipment. This paper focuses on fusing sensor data from Wi-Fi signals with computer vision methodologies in order to localize a user's mobile device with sub-meter accuracy. In particular, our algorithm detects doors in an office building automatically to calculate a user's position aided by WiFi signals.