Skip to content. | Skip to navigation

Personal tools


You are here: Home / Demos / inEvent navigation graph

inEvent navigation graph

The inEvent navigation graph is a prototype which demonstrates, over a small multimedia repository of lectures, how audio-visual processing can be used to segment lectures, and to derive features that are used to compute similarity between lectures and between segments. The most similar segments are displayed for the segments of the lecture in focus and can be used to view additional material related to a fragment of interest.Navigation graph

To view the demo, please use the following link: inEvent navigation graph

The demo allows users to visualize a lecture as a series of segments represented by keyword clouds or key frames (click to switch between the two), with relations to other similar lectures and segments. Segmentation was performed using a multi-factor algorithm which takes advantage of the automatic speech recognition to perform word-based segmentation, merging the results with detection from the video of actions such as writing on the blackboard. The similarity across segments and lectures is computed using a content-based recommendation algorithm. Overall, the graph-based representation of segment similarity appears to be a promising and cost-effective approach to navigating lecture databases.

This prototype has been designed for the ACM Multimedia 2013 Grand Challenge on Lecture Segmentation and Visualization, over the data set provided by the organizers from Videolectures.NET (which is the copyright holder for the audio-visual data). In addition to inEvent, the work was supported by the AROLES project of the Swiss National Science Foundation (51NF40-144627). If you wish to refer to this prototype, please refer to the following paper, which also contains a description of the underlying system:

Bhatt C., Popescu-Belis A., Habibi M., Ingram S., McInnes F., Masneri S., Pappas N. and Schreer O., “Multi-factor Segmentation for Topic Visualization and Recommendation: the MUST-VIS System”, Proceedings of ACM Multimedia 2013, Grand Challenge Solutions, Barcelona.