Advertisement

Semi-supervised domain adaptation for pedestrian detection in video surveillance based on maximum independence assumption

  • Ghazaleh Shojaei
  • Farbod RazzaziEmail author
Regular Paper
  • 7 Downloads

Abstract

In this paper, two domain adaptation approaches are utilized in a pedestrian detection application that is one of the most interesting and widely used fields in machine vision. In cases where the distributions of training and test data are different, the performance of pedestrian detection algorithms drops significantly. In this paper, employing two methods, namely transfer component analysis (TCA) and maximum independence domain adaptation (MIDA), the source and target domain data are represented in a new space where the distributions of two domains are more similar to each other, while the local geometry of data is preserved. Thereby, the classifier trained in the original space can be applied to the target data after the transformation. The experimental results of the proposed approach obtained on INRIA train dataset and CUHK test dataset show 82% about relative reduction in the classification error in the case of using TCA and about 84% in the case of using MIDA, compared to the baseline method where no domain adaptation method has been applied.

Keywords

Domain adaptation Pedestrian detection Transfer learning Semi-supervised learning Transfer component analysis Maximum independence domain adaptation 

Notes

References

  1. 1.
    Sun D, Watada J (2015) Detecting pedestrians and vehicles in traffic scene based on boosted HOG features and SVM. In: IEEE 9th International symposium on intelligent signal processing, WISP, pp 1–4Google Scholar
  2. 2.
    Dalal N, Triggs B (2005) Histograms of oriented gradients for human detection. In: IEEE computer society conference on computer vision and pattern recognition, CVPR, vol 1, pp 886–893Google Scholar
  3. 3.
    Felzenszwalb PF, Girshick RB, McAllester D (2010) Cascade object detection with deformable part models. In: IEEE conference on computer vision and pattern recognition, CVPR, pp 2241–2248Google Scholar
  4. 4.
    Dollár P, Tu Z, Perona, P et al (2009) Integral channel features. BMVC http://authors.library.caltech.edu
  5. 5.
    Felzenszwalb P, McAllester D, Ramanan D (2008) A discriminatively trained, multiscale deformable part model. In: IEEE conference on computer vision and pattern recognition (CVPR), pp 1–8Google Scholar
  6. 6.
    Dollár P, Belongie SJ, Perona P (2010) The fastest pedestrian detector in the west. In: Proceedings of the British machine vision conference, pp 68.1–68.11Google Scholar
  7. 7.
    Dollár P, Appel R, Belongie S et al (2014) Fast feature pyramids for object detection. IEEE Trans Pattern Anal Mach Intell 36(8):1532–1545CrossRefGoogle Scholar
  8. 8.
    Rosenberg C, Hebert M, Schneiderman H (2005) Semi-supervised self-training of object detection models. WACV/MOTION, (2)Google Scholar
  9. 9.
    Levin A, Viola PA, Freund Y (2003) Unsupervised improvement of visual detectors using co-training. In: IEEE international conference on computer vision, vol 1, pp 626–633Google Scholar
  10. 10.
    Javed O, Ali S, Shah M (2005) Online detection and classification of moving objects using progressively improving detectors. In: IEEE conference on computer vision and pattern recognition, CVPR, vol 1, pp 696–701Google Scholar
  11. 11.
    Yang J, Yan R, Hauptmann AG (2007) Cross-domain video concept detection using adaptive SVMs. In: Proceedings of 15th ACM international conference on multimedia, pp 188–197Google Scholar
  12. 12.
    Qi GJ, Aggarwal C, Huang T (2011) Towards semantic knowledge propagation from text corpus to web images. In: Proceedings of conference world wide web, pp 297–306Google Scholar
  13. 13.
    Duan L, Tsang IW, Xu D et al (2009) Domain transfer SVM for video concept detection. In: IEEE conference on computer vision and pattern recognition, pp 1375–1381Google Scholar
  14. 14.
    Pang J, Huang Q, Yan S et al (2011) Transferring boosted detectors towards viewpoint and scene adaptiveness. IEEE Trans Image Process 20(5):1388–1400MathSciNetCrossRefGoogle Scholar
  15. 15.
    Roth PM, Sternig S, Grabner H et al (2009) Classifier grids for robust adaptive object detection. In: Computer vision and pattern recognition, CVPR, IEEE Conference pp 2727–2734Google Scholar
  16. 16.
    Roth PM, Grabner H, Bischof H, et al (2005) On-line conservative learning for person detection. IEEE International Workshop on Visual Surveillance and Performance Evaluation of Tracking and Surveillance: 223-230Google Scholar
  17. 17.
    All K, Hasler D, Fleuret F (2011) Flow boost appearance learning from sparsely annotated video. In: IEEE conference on computer vision and pattern recognition, CVPR, pp 1433–1440Google Scholar
  18. 18.
    Jain V, Learned-Miller E (2011) Online domain adaptation of a pre-trained cascade of classifiers. In: IEEE conference on computer vision and pattern recognition, CVPR, pp 577–584Google Scholar
  19. 19.
    Rosenberg C, Hebert M, Schneiderman H (2005) Semi-supervised self-training of object detection models. In: 7th IEEE workshops on applications of computer vision (WACV/MOTION’05),vol 1, pp 29–36Google Scholar
  20. 20.
    Levin A, Viola PA, Freund Y (2003) Unsupervised improvement of visual detectors using co-training. In: Proceedings of the 9th IEEE international conference on computer vision, vol 1, pp 626–633Google Scholar
  21. 21.
    Wu B, Nevatia R (2007) Improving part based object detection by unsupervised, online boosting. In: IEEE conference on computer vision and pattern recognition, pp 1–8Google Scholar
  22. 22.
    Wang X, Wang M, Li W (2014) Scene-specific pedestrian detection for static video surveillance. IEEE Trans Pattern Anal Mach Intell 36(2):361–374MathSciNetCrossRefGoogle Scholar
  23. 23.
    Mao Y, Yin Z (2015) Training a scene-specific pedestrian detector using tracklets. In: IEEE winter conference on computer vision (WACV), pp 170–176Google Scholar
  24. 24.
    Maâmatou H, Chateau T, Gazzah S et al (2016) Sequential Monte Carlo filter based on multiple strategies for a scene specialization classifier. EURASIP J Image Video Process 1:40CrossRefGoogle Scholar
  25. 25.
    Pan SJ, Tsang IW, Kwok JT et al (2011) Domain adaptation via transfer component analysis. IEEE Trans Neural Netw 22(2):199–210CrossRefGoogle Scholar
  26. 26.
    Shi Y, Sha F (2012) Information-theoretical learning of discriminative clusters for unsupervised domain adaptation. In: Proceedings of the 29th international conference on machine learning, pp 1275–1282Google Scholar
  27. 27.
    Shao M, Kit D, Fu Y (2014) Generalized transfer subspace learning through low-rank constraint. Int J Comput Vision 109(1–2):74–93MathSciNetCrossRefGoogle Scholar
  28. 28.
    Pan SJ, Kwok JT, Yang Q (2008) Transfer learning via dimensionality reduction. In: Proceedings of the 23rd national conference on artificial intelligence, pp 677–682Google Scholar
  29. 29.
    Lanckriet GR, Cristianini N, Bartlett P et al (2004) Learning the kernel matrix with semidefinite programming. Mach Learn Res 5:27–72MathSciNetzbMATHGoogle Scholar
  30. 30.
    Song L, Smola A, Gretton A, Bedo J et al (2012) Feature selection via dependence maximization. Mach Learn Res 13:1393–1434MathSciNetzbMATHGoogle Scholar
  31. 31.
    Gretton A, Bousquet O, Smola A, et al (2005) Measuring statistical dependence with Hilbert-Schmidt norms. In: Conference algorithmic learning theory, Berlin, Heidelberg, pp 63–77Google Scholar
  32. 32.
    Yan K, Kou L, Zhang D (2018) Learning domain-invariant subspace using domain features and independence maximization. IEEE Trans Cybern 48(1):288–299CrossRefGoogle Scholar
  33. 33.
    Schölkopf B, Smola A, Müller KR (1998) Nonlinear component analysis as a kernel eigenvalue problem. Neural Comput 10(5):1299–1319CrossRefGoogle Scholar
  34. 34.
    Horn BK, Schunck BG (1981) Determining optical flow. Artif Intell 17(1–3):185–203CrossRefGoogle Scholar
  35. 35.
    Beauchemin SS, Barron JL (1995) The computation of optical flow. ACM Comput Surv (CSUR) 27(3):433–466CrossRefGoogle Scholar

Copyright information

© Springer-Verlag London Ltd., part of Springer Nature 2019

Authors and Affiliations

  1. 1.Department of Electrical and Computer Engineering Science and Research BranchIslamic Azad UniversityTehranIran

Personalised recommendations