1	Music Video Summarization Edward Wijaya Hendra Setiawan
2	Contents Multimedia Mining Multimedia Mining for Digital Library Music Video Summarization Conclusion
3	Multimedia Mining Summarization cast as Content Mining Problem – different with typical data mining Multimedia Mining – Pattern Discovery
4	Multimedia Mining for Digital Library Growing number of multimedia material in Digital Library catered by Content-based Retrieval Good summary is good candidate to facilitate user’s relevant judgement Good summary serve better indexing Accelerate the information seeking process (HCI aspects in Digital Libraries) for multimedia materials.
5	Video Summarization Strong visual-audio synchronization Audio-centric Visual-centric Weak visual-audio synchronization Summarize each track Alignment process
6	Music Video Summarization Weak synchronization Audio part is the salient part Goal : Provides a natural and effective audio-visual content overview Maximized the coverage for both audio and visual contents without having to sacrifice either of them [Shao Xi, et.al. Automatically Generating Summaries for Musical Video, 2003]
7	Music Summarization Problem : How to generate a concise and informative content that best summarize the original content Challenge : Featureless sequences of bytes (Feature Extraction is very important) →MPEG-7 standard Need content understanding Good candidate : Non-trivial repetitive pattern
8	Visual Summarization Challenges : Require overall understanding of the content Problem : Identify duplicates and redundancies in a video sequences (Easy) Finding the summary that has minimum redundancy
9	Visual and Audio Alignment Criteria : Smooth and Natural summary Maximize the coverage for both audio and visual content
10	Conclusion Summarization as Content Mining problem Multimedia Mining for Digital Library Music Video Summarization