Automatic object annotation for weakly labelled images/videos has attracted great research interests. In the literature, the idea of negative mining has been proposed for the task. Following existing works, the authors start with image/video over-segmentation. With the assumption that the noisy segments in the concept images and the strongly labelled non-concept segments are drawn from the same distribution, the authors plan to estimate the non-concept distribution and apply it to the ambiguous segments to generate a concept ranking. Although this idea was proposed in existing work and was shown ineffective when combined with a naive kernel density estimation strategy, in this study, the authors explore improved density estimation techniques for the ranking and propose a kernel regression model whose parameters are estimated by a maximum likelihood estimation. Experimental results validate the effectiveness of their method.

References

1. 1)
  - 17. Brox, T., Malik, J.: ‘Object segmentation by long term analysis of point trajectories’. European Conf. on Computer Vision, Heraklion, Crete, Greece, 2010, pp. 282–295.
2. 2)
  - 15. Arbelaez, P., Maire, M., Fowlkes, C., et al: ‘Contour detection and hierarchical image segmentation’, IEEE Trans. Pattern Anal. Mach. Intell., 2011, 33, (5), pp. 898–916.
3. 3)
  - 5. Zhu, J., Mao, J., Yuille, A.L.: ‘Learning from weakly supervised data by the expectation loss SVM (e-svm) algorithm’. Advances in Neural Information Processing Systems 27: Annual Conf. on Neural Information Processing Systems 2014, Montreal, QC, Canada, 2014, pp. 1125–1133.
4. 4)
  - 14. Felzenszwalb, P.F., Huttenlocher, D.P.: ‘Efficient graph-based image segmentation’, Int. J. Comput. Vis., 2004, 59, (2), pp. 167–181.
5. 5)
  - 1. Nguyen, M.H., Torresani, L., De la Torre, L., et al: ‘Weakly supervised discriminative localization and classification: a joint learning process’. IEEE Int. Conf. on Computer Vision, Kyoto, Japan, 2009, pp. 1925–1932.
6. 6)
  - 10. United States , 2013, pp. 2483–2490.
7. 7)
  - 19. Xu, C., Xiong, C., Corso, J.J.: ‘Streaming hierarchical video segmentation’. European Conf. on Computer Vision, Florence, Italy, 2012, pp. 626–639.
8. 8)
  - 22. Vedaldi, A., Fulkerson, B.: ‘VLFeat: An open and portable library of computer vision algorithms’, 2008. Accessed date: 2015, available at http://www.vlfeat.org/.
9. 9)
  - 8. Wang, S., Wang, Y.: ‘Weakly supervised semantic segmentation with a multiscale model’, IEEE Signal Process. Lett., 2015, 22, (3), pp. 308–312.
10. 10)
  - 23. Lowe, D.G.: ‘Distinctive image features from scale-invariant keypoints’, Int. J. Comput. Vis., 2004, 60, (2), pp. 91–110.
11. 11)
  - 11. Fu, Z., Robles-Kelly, A., Zhou, J.: ‘MILIS: multiple instance learning with instance selection’, IEEE Trans. Pattern Anal. Mach. Intell., 2011, 33, (5), pp. 958–977.
12. 12)
  - 9. Zhao, J., Wang, L., Cabral, R., et al: ‘Feature and region selection for visual learning’, IEEE Trans. Image Process., 2016, 25, pp. 1084–1094.
13. 13)
  - 18. Grundmann, M., Kwatra, V., Han, M., et al: ‘Efficient hierarchical graph-based video segmentation’. IEEE Conf. on Computer Vision and Pattern Recognition, San Francisco, CA, United States, 2010, pp. 2141–2148.
14. 14)
  - 3. Siva, P., Russell, C., Xiang, T.: ‘In defence of negative mining for annotating weakly labelled data’. European Conf. on Computer Vision, Florence, Italy, 2012, pp. 594–608.
15. 15)
  - 13. Chen, Y., Bi, J., Wang, J.Z.: ‘Miles: multiple-instance learning via embedded instance selection’, IEEE Trans. Pattern Anal. Mach. Intell., 2006, 28, (12), pp. 1931–1947.
16. 16)
  - 21. Ji, P., Zhao, N., Hao, S., et al: ‘Automatic image annotation by semi-supervised manifold kernel density estimation’, Inf. Sci., 2014, 281, pp. 648–660.
17. 17)
  - 20. Behmo, R., Marcombes, P., Dalalyan, A.S., et al: ‘Towards optimal naïve Bayes nearest neighbor’. European Conf. on Computer Vision, Heraklion, Crete, Greece, 2010, pp. 171–184.
18. 18)
  - 6. Bilen, H., Pedersoli, M., Tuytelaars, T.: ‘Weakly supervised object detection with convex clustering’. IEEE Conf. on Computer Vision and Pattern Recognition, Boston, MA, United States, pp. 1081–1089.
19. 19)
  - 4. Cinbis, R.G., Verbeek, J.J., Schmid, C.: ‘Multi-fold MIL training for weakly supervised object localization’. 2014 IEEE Conf. on Computer Vision and Pattern Recognition, Columbus, OH, United States, 2014, pp. 2409–2416.
20. 20)
  - 2. Siva, P., Xiang, T.: ‘Weakly supervised action detection’. British Machine Vision Conf., Dundee, United kingdom, 2011, pp. 1–11.
21. 21)
  - 16. Achanta, R., Shaji, A., Smith, K., et al: ‘SLIC superpixels compared to state-of-the-art superpixel methods’, IEEE Trans. Pattern Anal. Mach. Intell., 2012, 34, (11), pp. 2274–2282.
22. 22)
  - 10. Tang, K.D., Sukthankar, R., Yagnik, J., et al: ‘Discriminative segment annotation in weakly labeled video’. IEEE Conf. on Computer Vision and Pattern Recognition, Portland, OR,.
23. 23)
  - 7. Zhou, B., Jagadeesh, V., Piramuthu, R.: ‘Concept learner: discovering visual concepts from weakly labeled image collections’. IEEE Conf. on Computer Vision and Pattern Recognition, Boston, MA, United States, 2015, pp. 1492–1500.
24. 24)
  - 12. Jiang, H.: ‘Weakly supervised learning for salient object detection using background images. CoRR abs/1501.07492, 2015.

Non-concept density estimation via kernel regression for concept ranking in weakly labelled data

References

Related content