Content-Based Video Retrieval in Historical Collections of the German Broadcasting Archive
This addresses the need for efficient content-based retrieval in cultural heritage archives, but it is incremental as it applies existing video analysis methods to a new dataset.
The paper tackles the problem of automatically searching historical video collections by developing a system for the German Broadcasting Archive, achieving evaluation on 2,500 hours of GDR television recordings.
The German Broadcasting Archive (DRA) maintains the cultural heritage of radio and television broadcasts of the former German Democratic Republic (GDR). The uniqueness and importance of the video material stimulates a large scientific interest in the video content. In this paper, we present an automatic video analysis and retrieval system for searching in historical collections of GDR television recordings. It consists of video analysis algorithms for shot boundary detection, concept classification, person recognition, text recognition and similarity search. The performance of the system is evaluated from a technical and an archival perspective on 2,500 hours of GDR television recordings.