Smart-Pockets

摘要: This work introduces Smart-Pockets, a new set of whole-body gesture recognition techniques that enables users to access their personal digital content efficiently for visualization on ambient displays. Smart-Pockets works by recognizing users' body-deictic gestures entailing pockets, which associations between specific pockets and anchored those has been managed priori. The pocket metaphor we explore in this links using physical containers (i.e., pockets) placed at convenient locations the user's body, have specifically devised over decades fashion design store carry people's belongings comfortably conveniently. Consequently, are fast, require absolutely no precision perform effectively, robustly recognized user-independent scenarios with training required from user display. Also, technique is flexible easily extensible other containers, such as bags hand-held objects, demonstrate form Smart-Containers. We evaluate accuracy several gestures, report +99% classification explicit segmentation. discuss kinematic performance Smart-Containers show average time 2.2s comparable production touch smart mobile devices much smaller than produce gestures. Beyond practical implications advancing knowledge gesture-based interface interactions, believe contributions introduced concept will also foster developments pointing community attention toward (i) more examination potential class i.e., body-deictics, (ii) how public displays, an important preliminary step before actual interaction, (iii) inspiring examine creative objects visualized HighlightsSmart-Pockets displays.The creates link content.Smart-Pockets employ type hybrid deictic gestures.Smart-Pockets actions fast (2.2s) (+99%).The extendable