Comparing descriptions of multimedia data for simplification and integration

In this paper, different description of a same audio-visual material are considered. A general formulation for their integration is proposed. The descriptions are considered to be compliant with the specifications of the MPEG--7 standard. The idea is to manipulate information readily available from the considered descriptions to reach an accurate integration result, without having to reprocess the multimedia material. An example is shown in the context of a video sequence separation into shots. The proposed solution enables an integration of two separate temporal segmentations of the video sequence, leading to a possible automatic comparison of shot cut detection algorithms. The method uses the dominant color Descriptor information associated to the two original descriptions to reach the integration result.