Academic
Publications
User Transparent Parallel Processing of the 2004 NIST TRECVID Data Set

User Transparent Parallel Processing of the 2004 NIST TRECVID Data Set,10.1109/IPDPS.2005.443,Frank J. Seinstra,Cees Snoek,Dennis Koelma,Jan-mark Geus

User Transparent Parallel Processing of the 2004 NIST TRECVID Data Set   (Citations: 13)
BibTex | RIS | RefWorks Download
The Parallel-Horus framework, developed at the Univer- sity of Amsterdam, is a unique software architecture that al - lows non-expert parallel programmers to develop fully se- quential multimedia applications for efficient execution o n homogeneous Beowulf-type commodity clusters. Previously obtained results for realistic, but relatively small-size d ap- plications have shown the feasibility of the Parallel-Horu s approach, with parallel performance consistently being found to be optimal with respect to the abstraction level of message passing programs. In this paper we discuss the most serious challenge Parallel-Horus has had to deal with so far: the processing of over 184 hours of video included in the 2004 NIST TRECVID evaluation, i.e. the de facto in- ternational standard benchmark for content-based video re - trieval. Our results and experiences confirm that Parallel- Horus is a very powerful support-tool for state-of-the-art research and applications in multimedia processing.
Cumulative Annual
View Publication
The following links allow you to view full publications. These links are maintained by other sources not affiliated with Microsoft Academic Search.
    • ...For the processing of the visual modality in the content link we have therefore applied the Parallel-Horus software architecture [15]...
    • ...While we estimate that the processing of the entire TRECVID data set would have taken over 250 days on the fastest sequential machine available to us, application of Parallel-Horus in combination with a distributed Beowulf cluster consisting of 200 dual 1-Ghz Pentium-III CPUs [3] reduced the processing time to less than 48 hours [15]...

    C. G. M. Snoeket al. The MediaMill TRECVID 2010 semantic video search engine

    • ...TRECVID is computationally intensive; for thorough analysis it easily requires about 16 s of processing per video frame on the fastest sequential machine at our disposal [31]...

    Ran Yanget al. Resource optimization in distributed real-time multimedia applications

    • ...Notably, Parallel-Horus was applied in recent NIST TRECVID benchmark evaluations for content-based video retrieval, playing a crucial role in achieving top-ranking results in a field of strong international competitors [9], [10]...

    Fangbin Liuet al. Adaptive Parallel Householder Bidiagonalization

    • ...Our “TRECVID” application is computationally intensive; for thorough analysis it easily requires about 1 6 seconds of processing per video frame on the fastest sequential machine at our disposal [15]...

    R. Yanget al. On the Optimization of Resource Utilization in Distributed Multimedia ...

    • ...For the remaining image processing effort we apply the Parallel-Horus software architecture [29]...
    • ...Application of Parallel-Horus, in combination with a distributed cluster consisting of 200 dual 1-GHz Pentium-III CPUs [6], reduced the processing time to less than 60 hours [29]...

    Cees G. M. Snoeket al. Semantic Video Indexing

Sort by: