Vision Based Analysis on Trajectories of Notes Representing Ideas Toward Workshop Summarization

抄録

In some workshops, the participants use sticky notes to represent their ideas and the notes are regarded as one of the outputs. When workshop analysts summarize a workshop afterwards, they compare those notes and recorded video and/or audio by hand. We hypothesize that spatial and sequential information of the notes represents how ideas are came up with and how discussions flow. Therefore, we propose a vision based system analyzes trajectories of such notes during workshops. This paper introduces our prototype system that analyzes trajectories of notes representing ideas. During the workshop, the system recognizes and tracks each note separately from camera observation. The recognition and tracking results are saved and used for further offline analysis. From preliminary experiments, we confirmed that the proposed system can potentially reduce the manual tasks.