Lessons learned from building a terabyte digital video library

Wactlar, Howard D.; Christel, Michael G.; Gong, Yi; Hauptmann, Alex

doi:10.1109/2.745722

Cited by 165 publications

(108 citation statements)

References 10 publications

Supporting

Mentioning

108

Contrasting

Order By: Relevance

“…Thus desired video archives may need to satisfy autonomy also, which tends to be lacking in the former attempts. CMU Informedia project [8] built the video archive system which satisfies autonomy for specific types of programs, namely, news and documentaries. But it does not meet diversity.…”

Section: Design Issuesmentioning

confidence: 99%

Mining Large-Scale Broadcast Video Archives Towards Inter-video Structuring

Katayama

Ide

et al. 2004

Advances in Multimedia Information Processing - PCM 2004

View full text Add to dashboard Cite

Abstract. The current computer technology enables us to build huge broadcast video archives which had been a future dream. Today, even the hard disk recorders on the market are capable of recording several hundred hours of broadcast video. It is naturally perceived that a huge amount of broadcast video would be a useful corpus for multimedia indexing and mining research. Based on this viewpoint, we designed and constructed a broadcast video archive system having sufficient capacity and functionality to serve as the testbed for indexing and mining research. The system can capture multiple channels (currently seven channels) all-day broadcast video streams simultaneously, up to 6000 hours, and programspecific broadcasts, currently a news program for more than three years so far. This paper discusses design and implementation issues of the video archive system and then introduces our research efforts utilizing the archives as huge multimedia corpora.

show abstract

Section: Design Issuesmentioning

confidence: 99%

Mining Large-Scale Broadcast Video Archives Towards Inter-video Structuring

Katayama

Ide

et al. 2004

Advances in Multimedia Information Processing - PCM 2004

View full text Add to dashboard Cite

show abstract

“…-Better image and video compression algorithms and methodologies (Wactlar et al, 1999;Talagala et al, 2000), in order to construct Terabyte visual databases of Natural History images and videos.…”

Section: Analysis Layermentioning

confidence: 99%

What can bioinformatics do for Natural History museums?

Becerra¹

2003

GRAELLSIA

View full text Add to dashboard Cite

We propose the founding of a Natural History bioinformatics framework, which would solve one of the main problems in Natural History: data which is scattered around in many incompatible systems (not only computer systems, but also paper ones). This framework consists of computer resources (hardware and software), methodologies that ease the circulation of data, and staff expert in dealing with computers, who will develop software solutions to the problems encountered by naturalists.This system is organized in three layers: acquisition, data and analysis. Each layer is described, and an account of the elements that constitute it given. Key words: Taxonomy, Systematic, Bioinformatics. RESUMEN ¿Cómo puede ayudar la bioinformática a los Museos de Historia Natural?Se presentan las bases de una estructura bioinformática para Historia Natural, que trata de resolver uno de los principales problemas en ésta: la presencia de datos distribuidos a lo largo de muchos sistemas incompatibles entre sí (y no sólo hablamos de sistemas informáticos, sino también en papel). Esta estructura se sustenta en recursos informáticos (en sus dos vertientes: hardware y software), en metodologías que permitan la fácil circulación de los datos, y personal experto en el uso de ordenadores que se encargue de desarrollar soluciones software a los problemas que plantean los naturalistas. Este sistema estaría organizado en tres capas: de adquisición, de datos y de análisis. Cada una de estas capas se describe, indicando los elementos que la componen.

show abstract

“…Analyzing current multimedia search approaches IBM (2008); Wactlar et al (1999) from the time to search point of view, we observe that, if machine learning techniques are used to extract high-level concepts, large training times are required. The manual annotation alternative Shneiderman et al (2006), although essential in many cases, cannot keep the pace with the growth rate of video productions.…”

Section: Introductionmentioning

confidence: 99%