Content-based access to video objects: Temporal Segmentation, visual summarization, and feature extraction | Synapse