150 likes | 280 Vues
This document outlines the challenges and solutions in the field of automatic analysis of multimedia content, particularly in the context of extensive surveillance data. It discusses the significant processing demands arising from large datasets such as CCTV footage and broadcast TV, highlighting the necessity for advanced parallel and distributed computing solutions. Key insights from research conducted at the University of Amsterdam's Intelligent Systems Lab emphasize user transparency in utilizing gigascale processing capabilities. Strategies for overcoming integration issues and enhancing system performance are also explored.
E N D
… and Application Research Plans Frank J. Seinstra MultimediaN (BSIK Project) Intelligent Systems Lab AmsterdamInformatics InstituteUniversity of Amsterdam(Prof. Arnold Smeulders)
automatic analysis? A Real Problem, part 1… • News Broadcast - September 21, 2005 • Police investigating over 80.000 (!) CCTV recordings • First match found no earlier than 2.5 months after July 7 attacks
Image/Video Content Analysis • Lots of research + benchmark evaluations: • PASCAL-VOC (10,000+ images), TRECVID (200+ hours of video) • A Problem of scale: • At least 30-50 hours of processing time per hour of video! • Beeld&Geluid => 20.000 hours of TV broadcasts per year • NASA => over 850 Gb of hyper-spectral image data per day • London Underground => over 120.000 years of processing … !!!
Since 1998: “Parallel-Horus” DAS-type Clusters High Performance Computing • Solution: Very large scale parallel and distributed computing • New Problem: Very complicated software Solution: tool to make parallel & distributed computing transparent to user User Wide-Area Grid Systems Seinstra et al.: IEEE Trans. Par. Dist. Syst., 13(7), July 2002IEEE Trans. Par. Dist. Syst., 15(10), October 2004Parallel Computing, 28(7-8), August 2002Concur. Comput.: Pract. Exp., 16(6), May 2004
Extensions for Distributed Computing • Wide-Area Multimedia Services: Parallel Horus Client Parallel Horus Server Parallel Horus Servers Parallel Horus Servers Parallel Horus Client • User transparency? • Abstractions & techniques? • Integration: parallel/distributed?
A Real Problem, part 2… + LambdaRAM ?? may be time-critical…!
Example: Object Recognition See also: http://www.science.uva.nl/~fjseins/aibo.html
Example: Object Recognition Demonstrated live (a.o.) at ECCV 2006, June 8-11, 2006, Graz, Austria
Performance / Speedup on DAS-2 Single cluster, client side speedup Four clusters, client side speedup • Recognition on single machine: +/- 30 seconds • Using multiple clusters: up to 10 frames per second • Insightful: even ‘distant’ clusters can be used effectively for close to ‘real-time’ recognition
Ok, robot dog is a funny/crazy toy application, but: • Best performer in TRECVID 2004 & TRECVID 2005 Snoek et al., IEEE Trans. Pattern Anal. Mach. Intell. in press, 2006 Results: applicability • Beneficial: • Performance gains largely obtained ‘for free’ • With Parallel-Horus we can build similar complex ‘Grid’ applications in a matter of hours…
Current & Future Work • Very Large-Scale Distributed Multimedia Computing: • Overcome practical annoyances: • Software portability, firewall circumvention, authentication, … • Optimization and efficiency: • Tolerant to dynamic Grid circumstances, … • Systematic integration of MM-domain-specific knowledge, … • Deal with non-trivial communication patterns: • Heavy intra- & inter-cluster communication, … • Reach the end users: • Programming models, execution scenarios, … • Collaboration with VU (Prof. Henri Bal) & GridLab • Ibis: www.cs.vu.nl/ibis/ • Grid Application Toolkit: www.gridlab.org
But most of all: DAS-3 MATTERS !!!… not only to ‘C’ …… but also to ‘I’ in ‘ASCI’ … Conclusions • Effective integration of results from two largely distinct research fields • Ease of programming => quick solutions • With DAS-3 / StarPlane we can start to take on much more complicated problems