If you see something, swipe towards it: crowdsourced event localization using smartphones

摘要: This paper presents iSee, a crowdsourced approach to detecting and localizing events in outdoor environments. Upon spotting an event, iSee user only needs swipe on her smartphone's touchscreen the direction of event. These swiping directions are often inaccurate so compass measurements. Moreover, swipes do not encode any notion how far event is located from user, neither GPS location accurate. Furthermore, multiple may occur simultaneously users explicitly indicate which they towards. Nonetheless, as more start contributing data, we show that our proposed system able quickly detect estimate locations events. We have implemented Android phones experimented real-world settings by planting virtual "events" campus asking volunteers seeing one. Results performs appreciably better than established triangulation clustering-based approaches, terms localization accuracy, detection coverage, robustness sensor noise.