Abstract
The feature signatures in connection with the adaptive distance measures have become a respected similarity model for effective multimedia retrieval. However, the efficiency of the model is still a challenging task because the adaptive distance measures have at least quadratic time complexity according to the number of tuples in feature signatures. In order to reduce the number of tuples in feature signatures, we introduce the scalable feature signatures, a new formal framework enabling definition of new methods based on agglomerative hierarchical clustering. We show the framework can be used to express nontrivial feature signature reduction techniques including also popular agglomerative hierarchical clustering techniques. We experimentally demonstrate our new feature signature reduction techniques can be used to implement order of magnitude faster yet effective filter distances approximating the original adaptive distance measures. We also show the filter distances using our new feature signature reduction techniques can compete or even outperform the filter distances based on the related feature signature reduction techniques.
Similar content being viewed by others
Notes
Position 〈x,y〉, color 〈L,a,b〉 and texture information 〈c o n t r a s t,e n t r o p y〉, \(\mathbb {F} \subseteq \mathbb {R}^{7}\)
The maximal component feature signatures, centroid linkage method and Ward’s method do not assume a total ordering of the tuples comprising also weights and the lexicographic ordering.
The SQFD parameter α=1.28 performed sufficiently well for the TWIC and Profimedia datasets. However, for the ALOI dataset the effectiveness of all the reduction methods could be improved by SQFD parameter α for small number of tuples and thus for ALOI we have run the experiments also for α∈{0.001,0.08,0.16,0.32,0.64,1.28}.
Profimedia dataset is already provided with a set of query objects.
We utilize an optimized version using precomputed self-similarity matrices.
More comprehensive and rigorous correlation analysis involves inner workings of the utilized adaptive distance measures which is out of the scope of this paper.
The employed GPU extractor uses adaptive k-means that removes too small tuples and merges too similar tuples to reduce the size of the resulting feature signatures.
References
Achanta R, Shaji A, Smith K, Lucchi A, Fua P, Süsstrunk S (2012) Slic superpixels compared to state-of-the-art superpixel methods. IEEE Trans Pattern Anal Mach Intell 34 (11):2274–2282
Assent I, Wichterich M, Meisen T, Seidl T (2008) Efficient similarity search using the earth mover’s distance for large multimedia databases. In: Proceedings of IEEE ICDE, pp 307–316
Beecks C, Kirchhoff S, Seidl T (2013) Signature matching distance for content-based image retrieval. In: Proceedings of ACM international conference on multimedia retrieval (ICMR 2013), Dallas, Texas, USA. ACM, New York, pp 41–48
Beecks C, Uysal MS, Seidl T (2010) Efficient k-nearest neighbor queries with the signature quadratic form distance. In: Proceedings of 4th international workshop on ranking in databases (DBRank 2010) in conjunction with IEEE 26th international conference on data engineering (ICDE 2010), Long Beach, California, USA. IEEE, Washington, pp 10–15
Beecks C, Uysal MS, Seidl T (2010) Signature quadratic form distance. In: Proceedings of ACM CIVR, pp 438–445
Beecks C, Uysal MS, Seidl T (2011) L2-signature quadratic form distance for efficient query processing in very large multimedia databases. In: Proceedings of 17th international conference on multimedia modeling (MMM 2011), Taipei, LNCS, vol 6523. Springer, Berlin, pp 381–391
Bosch A, Zisserman A, Munoz X (2007) Representing shape with a spatial pyramid kernel. In: Proceedings of the 6th ACM international conference on image and video retrieval, CIVR ’07. ACM, New York, pp 401–408
Budikova P, Batko M, Zezula P (2011) Evaluation platform for content-based image retrieval systems. In: Proceedings of the 15th international conference on theory and practice of digital libraries: research and advanced technology for digital libraries, TPDL’11. Springer-Verlag, Berlin Heidelberg, pp 130–142
Chang S-F, Sikora T, Purl A (2001) Overview of the mpeg-7 standard. IEEE Trans Circ Syst Video Technol 11 (6):688–695
Chatzichristofis S, Boutalis Y, Lux M (2009) Selection of the proper compact composite descriptor for improving content based image retrieval. In: Proceedings of the 6th IASTED international conference, vol 134643, p 064
Chatzichristofis SA, Boutalis YS (2008) Cedd: C6olor and edge directivity descriptor: a compact descriptor for image indexing and retrieval. In: Proceedings of the 6th international conference on computer vision systems, ICVS’08. Springer-Verlag, Berlin Heidelberg, pp 312–322
Chatzichristofis SA, Boutalis YS (2008) Fcth: Fuzzy color and texture histogram - a low level feature for accurate image retrieval. In: Proceedings of the 2008 9th international workshop on image analysis for multimedia interactive services, WIAMIS’08. IEEE Computer Society, Washington, pp 191–196
Datta R, Joshi D, Li J, Wang JZ (2008) Image retrieval: ideas, influences, and trends of the new age. ACM Comput Surv 40 (2):5:1–5:60
Davidson I, Ravi SS (2009) Using instance-level constraints in agglomerative hierarchical clustering: Theoretical and empirical results. Data Min Knowl Discov 18 (2):257–282
Deselaers T, Keysers D, Ney H (2008) Features for image retrieval: an experimental comparison. Inf Retr 11 (2):77–107
Do T-T, Kijak E, Furon T, Amsaleg L (2010) Deluding image recognition in sift-based cbir systems. In: Proceedings of the 2nd ACM workshop on multimedia in forensics, security and intelligence, MiFor ’10. ACM, New York, pp 7–12
Everitt B, Landau S, Leese M, Stahl D (2011) Cluster analysis, 5th edn. Wiley, New York
Geusebroek J-M, Burghouts GJ, Smeulders AWM (2005) The amsterdam library of object images. IJCV 61 (1):103–112
Gordon AD (1987) A review of hierarchical classification. J R Stat Soc Series A (General) 150 (2):119–137
Hetland ML, Skopal T, Lokoč J, Beecks C (2013) Ptolemaic access methods: challenging the reign of the metric space model. Inf Syst 38 (7):989–1006
Huang J, Kumar SR, Mitra M, Zhu W-J, Zabih R (1997) Image indexing using color correlograms. In: Proceedings of the 1997 conference on computer vision and pattern recognition (CVPR ’97). IEEE Computer Society, Washington, p 762
Huttenlocher DP, Klanderman GA, Kl GA, Rucklidge WJ (1993) Comparing images using the Hausdorff distance. IEEE Trans Pattern Anal Mach Intell 15:850–863
Jegou H, Douze M, Schmid C (2008) Hamming embedding and weak geometric consistency for large scale image search. In: Proceedings of the 10th European conference on computer vision: Part I, ECCV ’08. Springer-Verlag, Berlin Heidelberg, pp 304–317
Jegou H, Perronnin F, Douze M, Sánchez J, Perez P, Schmid C (2012) Aggregating local image descriptors into compact codes. IEEE Trans Pattern Anal Mach Intell 34 (9):1704–1716
Kruliš M, Lokoč J, Skopal T (2013) Efficient extraction of feature signatures using multi-gpu architecture. In: MMM (2), LNCS, vol 7733. Springer Berlin Heidelberg, pp 446–456
Lance GN, Williams WT (1967) A general theory of classificatory sorting strategies: 1. hierarchical systems. The Comput J 9 (4):373–380
Lokoč J, (2014) Approximating the signature quadratic form distance using scalable feature signatures. In: Proceedings of 20th international conference on MultiMedia Modeling (MMM), volume 8325 of lecture notes in computer science. Springer International Publishing, pp 86–97
Lokoč J, Blažek A, Skopal T (2014) Signature-based video browser. In: Gurrin C, Hopfgartner F, Hurst W, Johansen H, Lee H, OConnor N (eds) MultiMedia modeling of lecture notes in computer science, vol 8326. Springer International Publishing, pp 415–418
Lokoč J, Novák D, Batko M, Skopal T (2012) Visual image search: feature signatures or/and global descriptors. In: Proceedings of the 5th international conference on similarity search and applications, SISAP’12. Springer-Verlag, Berlin Heidelberg, pp 177–191
Lux M (2013) Lire: Open source image retrieval in java. In: Proceedings of the 21st ACM international conference on multimedia, MM ’13. ACM, New York, pp 843–846
Mikolajczyk K, Schmid C (2005) A performance evaluation of local descriptors. IEEE Trans Pattern Anal Mach Intell 27 (10):1615–1630
MPEG-7 (2002) Multimedia content description interfaces. Part 3: Visual. ISO/IEC 15938–3:2002
Park B, Lee K, Lee S (2006) A new similarity measure for random signatures: perceptually modified hausdorff distance. In: Blanc-Talon J, Philips W, Popescu D, Scheunders P (eds) Advanced concepts for intelligent vision systems of lecture notes in computer science, vol 4179. Springer Berlin Heidelberg, pp 990–1001
Perronnin F, Liu Y, Sánchez J, Poirier H (2010) Large-scale image retrieval with compressed fisher vectors. In: IEEE conference on computer vision and pattern recognition (CVPR), pp 3384– 3391
Rubner Y, Tomasi C (2001) Perceptual metrics for image database navigation. Kluwer Academic Publishers, Norwell
Rubner Y, Tomasi C, Guibas LJ (2000) The earth mover’s distance as a metric for image retrieval. Int J Comput Vision 40 (2):99–121
Sivic J, Zisserman A (2003) Video google: a text retrieval approach to object matching in videos. In: Proceedings of the Ninth IEEE international conference on computer vision, ICCV ’03, vol 2. IEEE Computer Society, Washington, pp 1470–1477
Skopal T, Lokoc J, D-cache B Bustos (2012) Universal distance cache for metric access methods. IEEE Trans Knowl Data Eng 24 (5):868–881
Wagstaff K, Cardie C (2000) Clustering with instance-level constraints. In: Proceedings of the Seventeenth international conference on machine learning, ICML ’00. Morgan Kaufmann Publishers Inc, San Francisco, pp 1103–1110
Wu L, Hoi SCH, Yu N (2010) Semantics-preserving bag-of-words models and applications. Trans Img Proc 19 (7):1908–1920
Zhao H, Qi Z (2010) Hierarchical agglomerative clustering with ordering constraints. In: Proceedings of the 2010 Third international conference on knowledge discovery and data mining, WKDD ’10. IEEE Computer Society, Washington, pp 195–199
Acknowledgments
This research has been supported by Czech Science Foundation project GAČR P202/12/P297. I would also like to thank prof. Mathias Lux for his help with experiments regarding LIRE framework and provided detailed descriptions of utilized LIRE feature histograms and distance measures.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Lokoč, J. Approximating adaptive distance measures using scalable feature signatures. Multimed Tools Appl 74, 11569–11594 (2015). https://doi.org/10.1007/s11042-014-2251-4
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-014-2251-4