Random Methods for Fast Exploration of Raw Video Material

In this work we propose a visualization and access method for exploring a database of raw video material, where the span of information stored can be huge and difficult to understand at a glance. Developed visual mining tool presents the overview of the hierarchically structured repository in a intuitive way and provides interactive navigational support for the user. Browsing through the un-annotated content is enabled by two complementary interactive methods, a direct selection and a random access, both supported by visual display of the preview nodes. User evaluation aims at demonstrating how the hierarchical random visualization assists the process of accessing and retrieving content relevant to the user.