MediaMixer is pleased to initiate again in 2014 an ACM Multimedia 2014 Grand Challenge on the topic of Temporal segmentation and annotation of lecture videos. This follows on from the successful challenge in 2013.
The challenge is driven by the fact that VideoLectures.NET mostly hosts lectures 1 to 1.5 h long linked with slides and enriched with transcripts, metadata and additional textual contents. With automatic temporal segmentation and annotation of the video fragments, the VideoLectures.NET portal would gain in efficiency of the video search engine and would be able to provide users with the option to search for sections within a video, as well as would be able to recommend similar content to its users. This year the challenge again requires challenge participants to develop tools for the automatic segmentation and annotation of videos that could then be implemented in VideoLectures.NET. Extending lest year’s challenge definition and evaluation methodology, this year we will provide the challenge participants not only with a set of lecture videos and associated multimedia materials (e.g. slides, transcripts) but also with ground truth temporal segmenatation of the videos; the challenge participants will be asked to also report results against this ground truth. The ground truth data will be generated via crowdsourcing, which will be organized in advance by the challenge organizers.
Submission deadline is June 29, 2014. Contact the Grand Challenge organisers now to participate!