Promoting temporal diversity in video search results

Authors:
Martin Halvey;Anuj Goyal;Robert Villa;Joemon M. Jose
Affiliations:
University of Glasgow, Glasgow, Scotland, United Kingdom;University of Glasgow, Glasgow, Scotland, United Kingdom;University of Glasgow, Glasgow, Scotland, United Kingdom;University of Glasgow, Glasgow, Scotland, United Kingdom
Venue:
RIAO '10 Adaptivity, Personalization and Fusion of Heterogeneous Information
Year:
2010

Citing 3
Cited 0

Diversity, Assortment, Dissimilarity, Variety: A Study of Diversity Measures Using Low Level Features for Video Retrieval

ECIR '09 Proceedings of the 31th European Conference on IR Research on Advances in Information Retrieval
Generic and Spatial Approaches to Image Search Results Diversification

ECIR '09 Proceedings of the 31th European Conference on IR Research on Advances in Information Retrieval
User variance and its impact on video retrieval benchmarking

Proceedings of the ACM International Conference on Image and Video Retrieval

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper a number of methods for re-ranking video search results in order to introduce temporal diversity are presented. The usefulness of these approaches is evaluated in comparison with similarity based measures and visual diversity re-ranking approaches for TRECVID 2007 topics. The results of this evaluation indicate our temporal re-ranking has a negligible impact on retrieval performance. One of the most successful of these approaches was then implemented in an interactive search system for the TRECVID 2008 interactive search topics and compared with similarity based methods and a visual diversity re-ranking method. The results indicate that promoting temporal diversity does not have an adverse effect on user performance and in some cases can increase user performance.