Browse by topic
Type of publication
Revisiting the K-means Algorithm for Fast Trajectory Segmentation. Proceedings of the 38th International Conference on Computer Graphics and Interactive Techniques (SIGGRAPH), 2011.Many data in Computer Science are sequentially-generated and thus they require a trajectory segmentation; for instance, data from motion sensors, video cameras, RFID tags, eye trackers, etc. Segmentation leads to simplify the structure of the data, so that original objects can be divided into smaller, more compact structures that are better tailored for storage and retrieval purposes. We introduce a really straightforward method to cope with the sequentiality of data that is accurate, robust, extremely fast, and specially suited for real-time applications, large datasets, and on-line learning scenarios. Although this work is devoted to the segmentation and summarization of motion based trajectories, we believe that the range of potential applications span well beyond the computer graphics domain.