Enabling the Fusion Of Local Sensitivity And Low Rank Factorization To Mitigate The Risk Of Over Fitting

A. Anuradha, A.V.V. Satish


We propose a novel locality sensitive low-rank model for picture label finishing, which approximates the worldwide nonlinear model with a gathering of neighbourhood direct models. To viably imbue the possibility of territory sensitivity, a simple and compelling pre-handling module is intended to learn appropriate portrayal for information parcel, and a worldwide accord regularizer is acquainted with alleviate the danger of over fitting. In the interim, low-rank framework factorization is utilized as nearby models, where the local geometry structures are saved for the low-dimensional portrayal of both labels and tests. Broad experimental assessments led on three datasets exhibit the viability and proficiency of the proposed strategy, where our technique outflanks past ones by a vast edge.


H.-F. Yu,P. Jain, and I. S. Dhillon, “Large-scale multi-label learning withmissing labels,” in Proc. 31st Int. Conf. Mach. Learn., 2014, pp. 593–601.

M. M. Kalayeh, H. Idrees, and M. Shah, “NMF-KNN: Image annotationusing weighted multi-view non-negative matrix factorization,” in Proc.IEEE Conf. Comput. Vis. Pattern Recog., Jun.2014, pp. 184–191.482 IEEE TRANSACTIONS ON MULTIMEDIA, VOL. 18, NO. 3, MARCH 2016

S. Feng, R. Manmatha, and V. Lavrenko, “Multiple Bernoulli relevancemodels for image and video annotation,” in Proc. IEEE Conf. Comput.Vis. Pattern Recog., Jun. 2004, vol. 2, pp. 1002–1009.

G. Carneiro,A. B. Chan, P. J. Moreno, and N. Vasconcelos, “Supervisedlearning of semantic classes for image annotation and retrieval,” IEEETrans. Pattern Anal. Mach. Intell., vol. 29, no. 3, pp. 394–410, Mar.2007.

D. M. Blei and M. I. Jordan, “Modeling annotated data,” in Proc. Int.ACM SIGIR Conf. Res. Develop. Inform. Retrieval, 2003, pp. 127–134.

D. Putthividhy, H. T. Attias, and S. S. Nagarajan, “Topic regression multimodallatent dirichlet allocation for image annotation,” in Proc. IEEEConf. Comput. Vis. Pattern Recog., Jun. 2010, pp. 3408–3415.

C. Yang,M. Dong, and J. Hua, “Region-based image annotation usingasymmetrical support vector machine-based multiple-instance learning,”in Proc. IEEE Conf. Comput. Vis. Pattern Recog., Jun. 2006, vol. 2,pp. 2057–2063.

A. Makadia,V. Pavlovic, and S. Kumar, “A new baseline for image annotation,”in Proc. Eur. Conf. Compute. Vis., 2008, vol. 5304, pp. 316–329.

M. Guillaumin, T. Mensink, J. Verbeek, and C. Schmid, “TagProp: Discriminativemetric learning in nearest neighbor models for image autoannotation,”in Proc. IEEE Int. Conf. Comput. Vis., Sep.–Oct. 2009,pp. 309–316.

Y.VermaandC. Jawahar,“Image annotationusing metric learningin semantic neighbourhoods,” in Proc. Eur. Conf. Comput. Vis., 2012,pp. 836–849.[11] K. Q. Weinberger and L. K. Saul, “Distance metric learning for largemargin nearest neighbor classification,” J. Mach. Learn. Res.,vol. 10, pp.207–244, 2009.

S. S.Bucak,R. Jin, and A. K. Jain, “Multi-label learning with incompleteclass assignments,” in Proc. IEEE Conf. Comput. Vis. Pattern Recog.,Jun. 2011, pp. 2801–2808.

Y.Verma and C. Jawahar, “Exploring SVM for image annotation in presenceof confusing labels,” in Proc. Brit. Mach. Vis. Conf., 2013. [14] H.-F.Yu, P. Jain, P. Kar, and I. S. Dhillon, “Large-scale multilabellearning with missing labels,” CoRR, 2013.[Online]. Available: http://arxiv.org/abs/1307.5101.

M.Chen, A. Zheng, and K. Weinberger, “Fast image tagging,” in Proc.Int. Conf. Mach. Learn., 2013, pp. 1274–1282.

Full Text: PDF [Full Text]


  • There are currently no refbacks.

Copyright © 2013, All rights reserved.| ijseat.com

Creative Commons License
International Journal of Science Engineering and Advance Technology is licensed under a Creative Commons Attribution 3.0 Unported License.Based on a work at IJSEat , Permissions beyond the scope of this license may be available at http://creativecommons.org/licenses/by/3.0/deed.en_GB.