Efficient Boosting-Based Active Learning for Specific Object Detection Problems

Authors: Thuy Thi Nguyen, Nguyen Dang Binh, Horst Bischof

Abstract:

In this work, we present a novel active learning approach for learning a visual object detection system. Our system is composed of an active learning mechanism as wrapper around a sub-algorithm which implement an online boosting-based learning object detector. In the core is a combination of a bootstrap procedure and a semi automatic learning process based on the online boosting procedure. The idea is to exploit the availability of classifier during learning to automatically label training samples and increasingly improves the classifier. This addresses the issue of reducing labeling effort meanwhile obtain better performance. In addition, we propose a verification process for further improvement of the classifier. The idea is to allow re-update on seen data during learning for stabilizing the detector. The main contribution of this empirical study is a demonstration that active learning based on an online boosting approach trained in this manner can achieve results comparable or even outperform a framework trained in conventional manner using much more labeling effort. Empirical experiments on challenging data set for specific object deteciton problems show the effectiveness of our approach.

Keywords: Computer vision, object detection, online boosting, active learning, labeling complexity.

Digital Object Identifier (DOI): doi.org/10.5281/zenodo.1073016

References:

[1] Y. Abramson and Y. Freund. SEmi-automatic VIsual LEarning (SEVILLE): Tutorial on active learning for visual object recognition. Technical report, UCSD, 2005.
[2] S. Agarwal, A. Awan, and D. Roth. Learning to detect objects in images via a sparse, part-based representation. Transactions on Pattern Analysis and Machine Intelligence, 26(11):1475-1490, 2004.
[3] L. Anton-Canalis and E. Sanchez-Nielsen. Hand posture dataset creation for gesture recognition. In International Conference on Computer Vision Theory and Applications (VISAPP06), 2006.
[4] H. Francke, J. Ruiz del Solar, and R. Verschae. Real-time hand gesture detection and recognition using boosted classifiers and active learning. pages 533-547, 2007.
[5] H. Grabner and H. Bischof. On-line boosting and vision. In Proceedings CVPR, volume 1, pages 260-267, New York, NY, USA, 2006. IEEE Computer Society.
[6] R. Hewitt and S. Belongie. Active learning in face recognition: Using tracking to build a face model. CVPR Workshop, pages 157-157, 2006.
[7] C. Huang, H. Ai, Y. Li, and S. Lao. Vector boosting for rotation invariant multi-view face detection. ICCV, 2005, IEEE, 1:446-453 Vol. 1, 2005.
[8] O. Javed, S. Ali, and M. Shah. Online detection and classification of moving objects using progressively improving detectors. In Proceedings CVPR, pages 695-700, San Diego, CA, USA, 2005. IEEE.
[9] M. Kolsch and M. Turk. Robust hand detection. Proceedings, FG04, IEEE, pages 614-619, 2004.
[10] A. Levin, P. Viola, and Y. Freund. Unsupervised improvement of visual detectors using co-training. In Proceedings ICCV, volume 2, pages 626- 633, Nice, France, 2003. IEEE.
[11] Tong Luo, K. Kramer, S. Samson, and A. Remsen. Active learning to recognize multiple types of plankton. Proceedings ICPR 2004, 3:478- 481 Vol.3, 2004.
[12] C. Monteleoni and M. Kaariainen. Practical online active learning for classification. CVPR -07, IEEE, pages 1-8, 2007.
[13] Thuy Thi Nguyen, Helmut Grabner, Horst Bischof, and Barbara Gruber. On-line boosting for car detection from aerial images. RIVF-07, IEEE, pages 87-95, 2007.
[14] A. Opelt, M. Fussenegger, A. Pinz, and P. Auer. Weak hypotheses and boosting for generic object detection and recognition. In ECCV, pages Vol II: 71-84, 2004.
[15] N. Oza and S. Russell. Online bagging and boosting. In Proceedings AI and Statistics, pages 105-112, Florida, USA, 2001. Morgan Kaufmann.
[16] M.-Tri Pham and T.-J. Cham. Online learning asymmetric boosted classifiers for object detection. In CVPR-07, IEEE, 2007.
[17] C. Rosenberg, M. Hebert, and H. Schneiderman. Semi-supervised selftraining of object detection models. In Proceedings WACV. IEEE, January 2005.
[18] P. Roth and H. Bischof. On-line learning a person model from video data. In Video Proc. CVPR, IEEE, 2006.
[19] P.M. Roth, H. Grabner, D. Skoˇcaj, H. Bischof, and A. Leonardis. Online conservative learning for person detection. In Proceedings VS-PETS, pages 223-230, Beijing, China, 2005. IEEE.
[20] K. Sung and P. Niyogi. A formulation for active learning with applications to object detection. Technical Report AIM-1438, 1995.
[21] A. Torralba, K.P. Murphy, and W.T. Freeman. Sharing features: Efficient boosting procedures for multiclass object detection. volume 2, pages 762-769, 2005.
[22] R. Verschae, J. Ruiz del Solar, and M. Correa. A unified learning framework for object detection and classification using nested cascades of boosted classifiers. Mach. Vision Appl., 19(2):85-103, 2008.
[23] P. Viola and M. Jones. Rapid object detection using a boosted cascade of simple features. In CVPR, volume 1, pages 511-518, Kauai, HI, USA, 2001. IEEE Computer Society.
[24] Paul A. Viola and Michael J. Jones. Robust real-time face detection. IJCV, 57:137-154, 2004.
[25] T. Yang, J. Li, Q. Pan, C. Zhao, and Y. Zhu. Active learning based pedestrian detection in real scenes. In ICPR, pages 904-907, Washington, DC, USA, 2006. IEEE.