Intelligent Assistant System for Video Production on Mobile Devices

01.08.2016 - 30.09.2018
Research funding project

Modern mobile devices such as smartphones enable even the most novice user to easily take high quality pictures. Along with technological advances in the field of image sensory, the integration of user-supporting analyzing and correcting techniques (e.g. auto-exposure and face recognition) contributed significantly to an improvement in quality of the created content. Compared to taking individual pictures, the filming of high-quality videos takes much more effort for amateurs, as they possess little to none domain-specific knowledge. This project - a cooperation of emotion3D, a company which specializes in software engineering and image processing (e.g. 3D Vision Systems) for both film and post production, together with the research group Interactive Media Systems TU Vienna and experts in the field of user-centered-design at Technikum Vienna GmbH - aims at developing algorithms and their implementation in software tools in order to assist users when filming on mobile devices such as smartphones or tablets. More specifically, users receive easily comprehensible and automatically generated suggestions during recording in order to improve video direction (e.g. storyboard based concept) and are subsequently provided with feedback. The technical realization of the PersonalFilmAssistant is carried out through respective video processing algorithms, which extract low level scene information such as movement of objects for a subsequent knowledge domain transfer to an expert level of video direction.

People

Project leader

Subproject managers

Project personnel

Institute

Grant funds

  • Wirtschaftsagentur Wien Ein Fonds der Stadt Wien (National) FORSCHUNG (thematic Calls) Vienna Business Agency (WAW) Call identifier Call Users in Focus 2016

Research focus

  • Media Informatics and Visual Computing: 100%

Keywords

GermanEnglish
Mobile Computingmobile computing
Benutzerschnittstellenuser interface
Human Pose Estimationhuman pose estimation
Filmfilm
Videoanalysevideo analysis

Publications