Rapid Generation of Event-Based Indexes for Personalized Video Digests

already proposed a construction method for personalized video digests based on generated indexes [7][5]. This paper presents a novel video indexing method for providing timely personalized video delivery services. Several previous research reports have dealt with automatic content-bused indexing. However, these systems require a lot of time to manually correct the indexes after preliminary automatic processing, making it dijjkult to use these indexes in practical services. In our method, reliable content-bused indexes can be uutomuticully generated by manually flagging predejined and easily recognizable events while watching the video (without rewinding). The generated indexes use temporal functions of signijicunce bused on some parameters predefined for each kind of flagged event. Therefore, we can reduce the time required for flagging and achieve rupid semi-manual video indexing. Finally, we confirm effectiveness of the indexes transformed into a subset of MPEG-7 descriptions in a practical service to uutomuticully construct personalized video digests. Recently several automatic content-based video indexing methods have been described. Tusch et al. [9] defined an integrated generic video indexing model based on low-level visual features and high-level semantic objects. Other approaches include semantic video indexing using video streams, audio streams, texture and closed-caption information [4] [6]. Automatic indexing methods, however, require a lot of time to manually correct the indexes after preliminary automatic processing. This makes it difficult to use them in practical services.