Parallel Implementation of Neighbourhood Repulsed Correlation Metric Learning

Parallel Implementation of Neighbourhood Repulsed Correlation Metric Learning

  IJETT-book-cover  International Journal of Engineering Trends and Technology (IJETT)          
  
© 2019 by IJETT Journal
Volume-67 Issue-4
Year of Publication : 2019
Authors : Saurav Rai, Pallav Kumar Baruah
DOI :  10.14445/22315381/IJETT-V67I4P209

Citation 

MLA Style: Saurav Rai, Pallav Kumar Baruah "Parallel Implementation of Neighbourhood Repulsed Correlation Metric Learning" International Journal of Engineering Trends and Technology 67.4 (2019): 38-41.

APA Style:Saurav Rai, Pallav Kumar Baruah (2019). Parallel Implementation of Neighbourhood Repulsed Correlation Metric Learning. International Journal of Engineering Trends and Technology, 67(4), 38-41.

Abstract
Metric learning has attracted a lot of attention in recent times due to its usefulness and benefits in the domains of Machine learning, Natural Language Processing, and Big data. It is used as a pre-processing step in many Machine learning algorithms. However, as the size of the input data increases, the amount of time used for learning the metric which captures the semantic nature of the data also increases. In this paper, we have implemented parallel (CUDA) version of one such metric learning algorithm called as Neighbourhood Repulsed Correlation Metric Learning (NRCML). To illustrate the applicability of this parallel implementation, we derive motivation from the experimental evaluation that shows our implementation has greatly reduced the training time of the original sequential implementation. The proposed method has shown improvement in the performance of about 100x and above

Reference
[1]. Panagiotis Moutafis, Mengjun Leng, and Ioannis A Kakadiaris, "An overview and empirical comparision of distance metric learning methods," IEEE transactions on cybernectics, vol 47, no. 3, pp 612-625, 2017.
[2]. Habin Yan, "Kinship verification using neighborhood repulsed correlation metric learning." Image Vision Comput., vol. 60, pp. 91-97, 2017
[3]. Bo Xiao, Xiaokang Yang , Yi Xu, and Hongyuan Zha, " Learning distance metric for regression by semidefinite programming with application to human age estimation." in Proceedings of the 17th ACM international conference on Multimedia. ACM ,2009 , pp. 451-460.
[4]. Min Tan Zhenfang Hu, Baoyuan Wang, Jieyi Zhao, and Yeuming Wang, " Robust object recognition via weakly supervised metric and template learning", Neurocomputing, vol. 181, pp. 96-117, 2016
[5]. HM Sajjad Hossain, Md Abdullah Al Hafiz Khan, and Nirmalya Roy, "Active learning enabled activity recogntion," Pervasive and Mobile Computing, vol. 38, pp. 312-330 , 2017
[6]. Honghong Yang, Shiru Qu, and Zunxin Zheng, "Visual tracking, via online discriminative multiple instance metric learning", Multimedia Tools and Applications, pp. 1-19, 2017
[7]. Jiwen Lu, Gang Wang, and Jie Zhou," Simulataneous feature and dictionary learning for image set based face recognition." IEEE Transactions on Image Processing, 2017.
[8]. Rui Zhao, Wanli Oyang, and Xiaogang Wang, " Person reidentification by saliency learning," IEEE transactions on pattern analysis and machine intelligence, vol. 39, no.2 pp. 356-370, 2017

Keywords
Metric Learning, CUDA, NRCML.