V2F is a distributed video processing system for bounded (i.e., stored) and unbounded (i.e., continuous) or real time video streams. Apache Flink applies a series of operators in a pipeline to transform a video stream into shots. These operators are replicated to work in parallel on Flink-managed computing nodes. The V2F deployment of the standard twin-comparison video segmentation method is more than 7 times faster than its non-parallel (i.e., sequential) implementation.
CITATION STYLE
Kastrinakis, D., & Petrakis, E. G. M. (2022). V2F: Real Time Video Segmentation with Apache Flink. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 13599 LNCS, pp. 153–164). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-20716-7_12
Mendeley helps you to discover research relevant for your work.