Web-scale System for Image Similarity Search: When the Dreams Are Coming True

NOVÁK, David, Michal BATKO and Pavel ZEZULA. Web-scale System for Image Similarity Search: When the Dreams Are Coming True. In Proceedings of the Sixth International Workshop on Content-Based Multimedia Indexing (CBMI 2008). London: IEEE. p. 446-453. ISBN 978-1-4244-2043-8. 2008.

Basic information

• Original name: Web-scale System for Image Similarity Search: When the Dreams Are Coming True
• Name in Czech: Rozsáhlý systém pro podobnostní vyhledávání v obrázcích: Když se sny začínají naplňovat
• Authors: NOVÁK, David (203 Czech Republic, guarantor, belonging to the institution), Michal BATKO (203 Czech Republic, belonging to the institution) and Pavel ZEZULA (203 Czech Republic).
• Edition: London, Proceedings of the Sixth International Workshop on Content-Based Multimedia Indexing (CBMI 2008), p. 446-453, 8 pp. 2008.
• Publisher: IEEE

Other information

• Original language: English
• Type of outcome: Proceedings paper
• Field of Study: 10201 Computer sciences, information science, bioinformatics
• Country of publisher: United Kingdom of Great Britain and Northern Ireland
• Confidentiality degree: is not subject to a state or trade secret
• Publication form: printed version "print"
• WWW: URL
• RIV identification code: RIV/00216224:14330/08:00024279
• Organization unit: Faculty of Informatics
• ISBN: 978-1-4244-2043-8
• UT WoS: 000258985800060
• Keywords in English: similarity search; content-based search; image search; large-scale search; distributed data structures
• Tags: content-based search, DISA, distributed data structures, image search, large-scale search, similarity search
• Tags: International impact, Reviewed

Abstract

Digital images have become a commodity which is searched on the Web as ordinarily as web pages. However, current large-scale engines search the images only on the basis of their annotations, while the content-based similarity systems do not seem to be ready for such scales. In this paper, we open the way to Web-scale image similarity search. We present a flexible system based on the metric space model and on the peer-to-peer paradigm. It uses M-Chord and M-Tree structures as its fundamental components and measures the image similarity by a combination of five MPEG-7 features. The system has been implemented including a graphical interface for online demonstrations and it currently indexes 10 million images crawled from the Web. We propose a novel strategy for approximate evaluation of similarity queries and we test its performance by a series of experiments. The results show that the system provides high-quality answers with response times around 0.5 second.

Abstract (in Czech)

V této práci otevíráme cestu k podobnostnímu vyhlédávání na obrázcích v rozsahu Webu. Prezentujeme univerzální a flexibilní systém založený na metrickém modelu dat a konceptu peer-to-peer. Systém byl implementován a v současnosti indexuje 10 milionů obrázků. Experimenty prokazují, že systém poskytuje velmi kvalitní výsledky s dobou odezvz okolo půl sekundy.

Links

• GD102/05/H050, research and development project: Name: Integrovaný přístup k výchově studentů DSP v oblasti paralelních a distribuovaných systémů

Investor: Czech Science Foundation, Integrated approach to education of PhD students in the area of parallel and distributed systems

• GP201/08/P507, research and development project: Name: Komplexní podobnostní dotazy nad rozsáhlými objemy dat

Investor: Czech Science Foundation, Complex similarity searching in very large data collections

• 1ET100300419, research and development project: Name: Inteligentní modely, algoritmy, metody a nástroje pro vytváření sémantického webu

Investor: Academy of Sciences of the Czech Republic, Intelligent Models, Algorithms, Methods and Tools for the Semantic Web (realization)