Special Issue Article
Content Based Analysis Improves Audiovisual Archieve Retrieval
Media professionals actively utilize audiovisual archives as a source for reusable material. Archives are struggling to reinvent themselves in the face of fully digital operations and growing user bases. Yet, surprisingly, very little has been done to examine how content-based video retrieval will affect the searches of professionals searching in the audiovisual archive.So the primary goal is to investigate how content-based video search which enhances the performance of traditional archive retrieval. The project complements the old, manual, descriptions of the images in the archive with new, automatically generated, labels. Then, this project measures the effect of combining them for queries typical of professionals searching an archive. The queries used present are not based on real-world queries, and generally no manually created metadata (which is often present in the real world) is included in the experiments. The project takes into account the information needs and retrieval data already present in the audiovisual archive, and demonstrate that retrieval performance can be significantly improved when content-based methods are applied to search.