Quantifying the Specificity of Near-duplicate Image Classification Functions

Richard Connor, Franco Alberto Cardillo

Abstract

There are many published methods for detecting similar and near-duplicate images. Here, we consider their use in the context of unsupervised near-duplicate detection, where the task is to find a (relatively small) near- duplicate intersection of two large candidate sets. Such scenarios are of particular importance in forensic near-duplicate detection. The essential properties of a such a function are: performance, sensitivity, and specificity. We show that, as collection sizes increase, then specificity becomes the most important of these, as without very high specificity huge numbers of false positive matches will be identified. This makes even very fast, highly sensitive methods completely useless. Until now, to our knowledge, no attempt has been made to measure the specificity of near-duplicate finders, or even to compare them with each other. Recently, a benchmark set of near-duplicate images has been established which allows such assessment by giving a near-duplicate ground truth over a large general image collection. Using this we establish a methodology for calculating specificity. A number of the most likely candidate functions are compared with each other and accurate measurement of sensitivity vs. specificity are given. We believe these are the first such figures be to calculated for any such function.

References

  1. Bober, M. (2001). Mpeg-7 visual shape descriptors. IEEE Transactions on circuits and systems for video technology, 11(6):716-719.
  2. Bolettieri, P., Esuli, A., Falchi, F., Lucchese, C., Perego, R., Piccioli, T., and Rabitti, F. (2009). Cophir: a test collection for content-based image retrieval. CoRR, abs/0905.4627.
  3. Chum, O., Philbin, J., Isard, M., and Zisserman, A. (2007). Scalable near identical image and shot detection. In Proceedings of the 6th ACM international conference on Image and video retrieval, pages 549-556. ACM.
  4. Connor, R. (2015). Mir-flickr near-duplicate data. mirflickr-near-duplicates.appspot.com.
  5. Connor, R., Cardillo, F., MacKenzie-Leigh, S., and Moss, R. (2015). Identification of mir-flickr near-duplicate images. In 10th International Conference on Computer Vision Theory and Applications.
  6. Connor, R. and Moss, R. (2012). A multivariate correlation distance for vector spaces. In Navarro, G. and Pestov, V., editors, Similarity Search and Applications, volume 7404 of Lecture Notes in Computer Science, pages 209-225. Springer Berlin Heidelberg.
  7. Connor, R., Simeoni, F., Iakovos, M., and Moss, R. (2011). A bounded distance metric for comparing tree structure. Inf. Syst., 36(4):748-764.
  8. Foo, J., Sinha, R., and Zobel, J. (2006). Discovery of image versions in large collections. In Cham, T.-J., Cai, J., Dorai, C., Rajan, D., Chua, T.-S., and Chia, L.-T., editors, Advances in Multimedia Modeling, volume 4352 of Lecture Notes in Computer Science, pages 433- 442. Springer Berlin Heidelberg.
  9. Huiskes, M. J. and Lew, M. S. (2008). The mir flickr retrieval evaluation. In MIR 7808: Proceedings of the 2008 ACM International Conference on Multimedia Information Retrieval, New York, NY, USA. ACM.
  10. Huiskes, M. J., Thomee, B., and Lew, M. S. (2010). New trends and ideas in visual concept detection: The MIR Flickr retrieval evaluation initiative. In MIR 7810: Proceedings of the 2010 ACM International Conference on Multimedia Information Retrieval, pages 527-536, New York, NY, USA. ACM.
  11. ISO-15938. Mpeg-7 multimedia content description interface.
  12. Jinda-Apiraksa, A., Vonikakis, V., and Winkler, S. (2013). California-nd: An annotated dataset for near-duplicate detection in personal photo collections. In Quality of Multimedia Experience (QoMEX), 2013 Fifth International Workshop on, pages 142-147. IEEE.
  13. Lin, J. (1991). Divergence measures based on the shannon entropy. Information Theory, IEEE Transactions on, 37(1):145-151.
  14. Niu, X.-m. and Jiao, Y.-h. (2008). An overview of perceptual hashing. Acta Electronica Sinica, 36(7):1405- 1411.
  15. Oliva, A. and Torralba, A. (2001). Modeling the shape of the scene: A holistic representation of the spatial envelope. International Journal of Computer Vision, 42(3):145-175.
  16. Ventura Royo, C. (2010). Image-based query by example using mpeg-7 visual descriptors.
  17. Vonikakis, V., Jinda-Apiraksa, A., and Winkler, S. (2014). Photocluster - a multi-clustering technique for nearduplicate detection in personal photo collections. In Proc. of the 9th International Conference on Computer Vision Theory and Applications, pages 153-161.
  18. Won, C. S., Park, D. K., and Park, S.-J. (2002). Efficient use of mpeg-7 edge histogram descriptor. Etri Journal, 24(1):23-30.
Download


Paper Citation


in Harvard Style

Connor R. and Cardillo F. (2016). Quantifying the Specificity of Near-duplicate Image Classification Functions . In Proceedings of the 11th Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4: VISAPP, (VISIGRAPP 2016) ISBN 978-989-758-175-5, pages 647-654. DOI: 10.5220/0005785406470654


in Bibtex Style

@conference{visapp16,
author={Richard Connor and Franco Alberto Cardillo},
title={Quantifying the Specificity of Near-duplicate Image Classification Functions},
booktitle={Proceedings of the 11th Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4: VISAPP, (VISIGRAPP 2016)},
year={2016},
pages={647-654},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0005785406470654},
isbn={978-989-758-175-5},
}


in EndNote Style

TY - CONF
JO - Proceedings of the 11th Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4: VISAPP, (VISIGRAPP 2016)
TI - Quantifying the Specificity of Near-duplicate Image Classification Functions
SN - 978-989-758-175-5
AU - Connor R.
AU - Cardillo F.
PY - 2016
SP - 647
EP - 654
DO - 10.5220/0005785406470654