The proliferation of smartphones with video recording and action cameras has empowered users to effortlessly capture everyday moments and activities in video format. This process often results into long video sequences that need to be edited down and summarized in order to be properly archived, shared within social media, or simply projected to family and friends. For those who do not possess video-editing skills, isolating the most appropriate moments is a difficult and time-consuming task. This work presents "ActionTrip", a system that provides gestural user interaction during video-capture, in order to seamlessly tag personal points of interest. In addition, the system captures contextual information about nearby landmarks and information from social networking systems. This combination of personal with social preferences has enabled a single-step final editing process, which presents to the user a small list of points interest (public landmarks and personal preferences), instead of the actual recorded video that might be very long to comprehend. The benefits of this system have been demonstrated through a small-scale case study during the visit and tour in a European city.

PDF DOI  

Deliyannis, I., Symianakis, S., and Chorianopoulos, K. 2018. ActionTrip: automating egocentric video production through gestural interaction and landmark awareness. Proceedings of the 22nd Pan-Hellenic Conference on Informatics, ACM, 127–132.   BibTeX