access icon free Joint attribute chain prediction for zero-shot learning

© The Institution of Engineering and Technology
Received 04/10/2017, Accepted 16/04/2018, Revised 15/01/2018, Published 26/04/2018

Zero-shot learning (ZSL) aims to classify the objects without any training samples. Attributes are used to transfer knowledge from the training set to testing one in ZSL. Most ZSL methods based on Direct Attribute Prediction (DAP) assume that attributes are independent of each other. In this study, the authors explore the relationship between attributes and propose Joint Attribute Chain Prediction (JACP). Attribute chains are introduced to represent the relations. Conditional probabilities of attributes are estimated orderly along the chain to calculate the joint posteriors of the testing classes without independence assumptions. To reduce the estimation error, attribute relation clustering algorithm is presented to group the long chain into some unrelated small chains. When the max length of chains is one, JACP is essentially identical with DAP. Experiments on three data sets for zero-shot problem demonstrate the classification accuracy and efficiency of the authors’ algorithm. The results show that mining attribute relations can greatly improve the performance of ZSL effectively.

Inspec keywords: probability; data mining; learning (artificial intelligence); image classification

Other keywords: classification accuracy; direct attribute prediction; attribute relation clustering algorithm; estimation error reduction; conditional probabilities; JACP; ZSL methods; joint attribute chain prediction; zero-shot learning; attribute relation mining

Subjects: Knowledge engineering techniques; Image recognition; Other topics in statistics; Computer vision and image processing techniques; Other topics in statistics

References

    1. 1)
      • 3. Jing, Z., Qiao, L., Pan, H., et al: ‘An overview of the configuration and manipulation of soft robotics for on-orbit servicing’, SCIENCE CHINA Inf. Sci., 2017, 60, (5), 050201.
    2. 2)
      • 13. Suzuki, M., Sato, H., Oyama, S., et al: ‘Image classification by transfer learning based on the predictive ability of each attribute’. Proc. the Int. MultiConf. of Engineers and Computer Scientists, Hong Kong, China, March 2014, vol. 1.
    3. 3)
      • 16. Akata, Z., Perronnin, F., Harchaoui, Z., et al: ‘Label-embedding for attribute-based classification’. Proc. Computer Vision and Pattern Recognition, Portland, USA, June 2013, pp. 819–826.
    4. 4)
      • 38. Lai, Z., Xu, Y., Chen, Q., et al: ‘Multilinear sparse principal component analysis’, IEEE Trans. Neural Netw. Learn. Syst., 2014, 25, (10), pp. 1942–1950.
    5. 5)
      • 21. Socher, R., Ganjoo, M., Manning, C.D., et al: ‘Zero-shot learning through cross-modal transfer’. Proc. NIPS, Lake Tahoe, USA, December 2013, pp. 935–943.
    6. 6)
      • 8. Wang, X., Ji, Q.: ‘A unified probabilistic approach modeling relationships between attributes and objects’. Proc. the IEEE Int. Conf. on Computer Vision, Sydney, Australia, December 2013, pp. 2120–2127.
    7. 7)
      • 27. Schapire, R.E., Singer, Y.: ‘Boostexter: a boosting-based system for text categorization’, Mach. Learn., 2000, 39, (2-3), pp. 135–168.
    8. 8)
      • 36. Patterson, G., Xu, C., Su, H., et al: ‘The sun attribute database: beyond categories for deeper scene understanding’, Int. J. Comput. Vis., 2014, 108, (1-2), pp. 59–81.
    9. 9)
      • 28. Elisseeff, A., Weston, J.: ‘A kernel method for multi-labelled classification’. Proc. NIPS, Vancouver, Canada, December 2002, pp. 681–687.
    10. 10)
      • 10. Lampert, C.H., Nickisch, H., Harmeling, S.: ‘Learning to detect unseen object classes by between-class attribute transfer’. Proc. Computer Vision and Pattern Recognition, Miami, USA, June 2009, pp. 951–958.
    11. 11)
      • 9. Cheng, W., Hüllermeier, E., Dembczynski, K.J.: ‘Bayes optimal multilabel classification via probabilistic classifier chains’. Proc. the 27th int. Conf. on Machine Learning, Haifa, Israel, June 2010, pp. 279–286.
    12. 12)
      • 26. Godbole, S., Sarawagi, S.: ‘Discriminative methods for multi-labeled classification’, Pacific-Asia Conf. on Knowledge Discovery and Data Mining, May 2004, pp. 22–30.
    13. 13)
      • 12. Kankuekul, P., Kawewong, A., Tangruamsub, S., et al: ‘Online incremental attribute-based zero-shot learning’. Proc. Computer Vision and Pattern Recognition, Providence, USA, June 2012, pp. 3657–3664.
    14. 14)
      • 23. Kodirov, E., Xiang, T., Fu, Z., et al: ‘Unsupervised domain adaptation for zero-shot learning’. Proc. Int. Conf. on Computer Vision, Santiago, Chile, December 2015, pp. 2452–2460.
    15. 15)
      • 7. Liu, M., Zhang, D., Chen, S.: ‘Attribute relation learning for zero-shot classification’, Neurocomputing, 2014, 139, pp. 34–46.
    16. 16)
      • 37. Qiao, L., Tuo, H., Fang, Z., et al: ‘Joint probability estimation of attribute chain for zero-shot learning’. Proc. ICIP, Phoenix, USA, September 2016, pp. 1863–1867.
    17. 17)
      • 22. Zhang, Z., Saligrama, V.: ‘Zero-shot learning via joint latent similarity embedding’. Proc. Computer Vision and Pattern Recognition, Las Vegas, USA, June 2016, pp. 6034–6042.
    18. 18)
      • 35. Farhadi, A., Endres, I., Hoiem, D., et al: ‘Describing objects by their attributes’. Proc. Computer Vision and Pattern Recognition, Miami, USA, June 2009, pp. 1778–1785.
    19. 19)
      • 15. Li, H., Li, D., Luo, X.: ‘Bap: bimodal attribute prediction for zero-shot image categorization’. Proc. the 22nd ACM Int. Conf. on Multimedia, Orlando, USA, November 2014, pp. 1013–1016.
    20. 20)
      • 25. Zhang, Z., Saligrama, V.: ‘Zero-shot recognition via structured prediction’. Proc. European Conf. on Computer Vision, Amsterdam, The Netherlands, October 2016, pp. 533–548.
    21. 21)
      • 39. Lai, Z., Mo, D., Wong, W.K., et al: ‘Robust discriminant regression for feature extraction’, IEEE Trans. Cybern., 2017, in press.
    22. 22)
      • 4. Zhang, W., Sun, F., Wu, H., et al: ‘A framework for the fusion of visual and tactile modalities for improving robot perception’, SCIENCE CHINA Inf. Sci., 2017, 60, (1), 012201.
    23. 23)
      • 31. Zaragoza, J.H., Sucar, L.E., Morales, E.F., et al: ‘Bayesian chain classifiers for multidimensional classification’. Proc. IJCAI, Barcelona, Spain, July 2011, vol. 11, pp. 2192–2197.
    24. 24)
      • 32. Bielza, C., Li, G., Larranaga, P.: ‘Multi-dimensional classification with Bayesian networks’, Int. J. Approx. Reason., 2011, 52, (6), pp. 705–727.
    25. 25)
      • 1. Lowe, D.G.: ‘Distinctive image features from scale-invariant keypoints’, Int. J. Comput. Vision, 2004, 60, (2), pp. 91–110.
    26. 26)
      • 11. Huang, S., Elhoseiny, M., Elgammal, A., et al: ‘Learning hypergraph-regularized attribute predictors’. Proc. Computer Vision and Pattern Recognition, Boston, USA, June 2015, pp. 409–417.
    27. 27)
      • 18. Fu, Z., Xiang, T., Kodirov, E., et al: ‘Zero-shot object recognition by semantic manifold distance’. Proc. Computer Vision and Pattern Recognition, Boston, USA, June 2015, pp. 2635–2644.
    28. 28)
      • 14. Jayaraman, D., Grauman, K.: ‘Zero-shot recognition with unreliable attributes’. Proc. NIPS, Montreal, Canada, December 2014, pp. 3464–3472.
    29. 29)
      • 30. Tsoumakas, G., Katakis, I., Vlahavas, I.: ‘Random k-label sets for multilabel classification’, IEEE Trans. Knowl. Data Eng., 2011, 23, (7), pp. 1079–1089.
    30. 30)
      • 20. Frome, A., Corrado, G.S., Shlens, J., et al: ‘Devise: A deep visual-semantic embedding model’. Proc. NIPS, Lake Tahoe, USA, December 2013, pp. 2121–2129.
    31. 31)
      • 19. Zhang, Z., Saligrama, V.: ‘Zero-shot learning via semantic similarity embedding’. Proc. Int. Conf. on Computer Vision, Santiago, Chile, December 2015, pp. 4166–4174.
    32. 32)
      • 5. Lampert, C.H., Nickisch, H., Harmeling, S.: ‘Attribute-based classification for zero-shot visual object categorization’, IEEE Trans. Pattern Anal. Mach. Intell., 2014, 36, (3), pp. 453–465.
    33. 33)
      • 17. Fu, Y., Hospedales, T.M., Xiang, T., et al: ‘Transductive multi-view embedding for zero-shot recognition and annotation’. Proc. European Conf. on Computer Vision, ZÃijrich, Switzerland, September 2014, pp. 584–599.
    34. 34)
      • 6. Wang, Y., Mori, G.: ‘A discriminative latent model of object classes and attributes’. Proc. European Conf. on Computer Vision, Hersonissos, Greece, September 2010, pp. 155–168.
    35. 35)
      • 34. Platt, J.: ‘Probabilistic outputs for support vector machines and comparisons to regularized likelihood methods’, Advances Large Margin Classifiers, 1999, 10, (3), pp. 61–74.
    36. 36)
      • 24. Fu, Y., Hospedales, T.M., Xiang, T., et al: ‘Transudative multi-view zero-shot learning’, IEEE Trans. Pattern Anal. Mach. Intell., 2015, 37, (11), pp. 2332–2345.
    37. 37)
      • 33. Dembszynski, K., Waegeman, W., Cheng, W., et al: ‘On label dependence in multilabel classification’. ICML Workshop, Haifa, Israel, June 2010.
    38. 38)
      • 29. Zhang, M.L., Zhou, Z.H.: ‘ML-KNN: A lazy learning approach to multi-label learning’, Pattern Recognit., 2007, 40, (7), pp. 2038–2048.
    39. 39)
      • 2. Dalal, N., Triggs, B.: ‘Histograms of oriented gradients for human detection’. Proc. Computer Vision and Pattern Recognition, San Diego, USA, June 2005, pp. 886–893.
http://iet.metastore.ingenta.com/content/journals/10.1049/iet-cvi.2017.0438
Loading

Related content

content/journals/10.1049/iet-cvi.2017.0438
pub_keyword,iet_inspecKeyword,pub_concept
6
6
Loading