Automatically Detecting Important Moments From Everyday Life Using a Mobile Device
This paper proposes a new method to detect important moments in the people lives. The authors' paper is motivated by the increase in the quantity of multimedia data, such as videos and photos, which are capturing life experiences into personal archives. Even though such media-rich data suggests visual processing to identify important moments, the oft-mentioned problem of the semantic gap means that users cannot automatically identify or retrieve important moments using visual processing techniques alone. Their approach utilises on-board sensors from mobile devices to automatically identify important moments, as they are happening.