{"references": ["Y. Abramson and Y. Freund. SEmi-automatic VIsual LEarning\n(SEVILLE): Tutorial on active learning for visual object recognition.\nTechnical report, UCSD, 2005.", "S. Agarwal, A. Awan, and D. Roth. Learning to detect objects in images\nvia a sparse, part-based representation. Transactions on Pattern Analysis\nand Machine Intelligence, 26(11):1475-1490, 2004.", "L. Anton-Canalis and E. Sanchez-Nielsen. Hand posture dataset creation\nfor gesture recognition. In International Conference on Computer Vision\nTheory and Applications (VISAPP06), 2006.", "H. Francke, J. Ruiz del Solar, and R. Verschae. Real-time hand gesture\ndetection and recognition using boosted classifiers and active learning.\npages 533-547, 2007.", "H. Grabner and H. Bischof. On-line boosting and vision. In Proceedings\nCVPR, volume 1, pages 260-267, New York, NY, USA, 2006. IEEE\nComputer Society.", "R. Hewitt and S. Belongie. Active learning in face recognition: Using\ntracking to build a face model. CVPR Workshop, pages 157-157, 2006.", "C. Huang, H. Ai, Y. Li, and S. Lao. Vector boosting for rotation invariant\nmulti-view face detection. ICCV, 2005, IEEE, 1:446-453 Vol. 1, 2005.", "O. Javed, S. Ali, and M. Shah. Online detection and classification of\nmoving objects using progressively improving detectors. In Proceedings\nCVPR, pages 695-700, San Diego, CA, USA, 2005. IEEE.", "M. Kolsch and M. Turk. Robust hand detection. Proceedings, FG04,\nIEEE, pages 614-619, 2004.\n[10] A. Levin, P. Viola, and Y. Freund. Unsupervised improvement of visual\ndetectors using co-training. In Proceedings ICCV, volume 2, pages 626-\n633, Nice, France, 2003. IEEE.\n[11] Tong Luo, K. Kramer, S. Samson, and A. Remsen. Active learning to\nrecognize multiple types of plankton. Proceedings ICPR 2004, 3:478-\n481 Vol.3, 2004.\n[12] C. Monteleoni and M. Kaariainen. Practical online active learning for\nclassification. CVPR -07, IEEE, pages 1-8, 2007.\n[13] Thuy Thi Nguyen, Helmut Grabner, Horst Bischof, and Barbara Gruber.\nOn-line boosting for car detection from aerial images. RIVF-07, IEEE,\npages 87-95, 2007.\n[14] A. Opelt, M. Fussenegger, A. Pinz, and P. Auer. Weak hypotheses and\nboosting for generic object detection and recognition. In ECCV, pages\nVol II: 71-84, 2004.\n[15] N. Oza and S. Russell. Online bagging and boosting. In Proceedings AI\nand Statistics, pages 105-112, Florida, USA, 2001. Morgan Kaufmann.\n[16] M.-Tri Pham and T.-J. Cham. Online learning asymmetric boosted\nclassifiers for object detection. In CVPR-07, IEEE, 2007.\n[17] C. Rosenberg, M. Hebert, and H. Schneiderman. Semi-supervised selftraining\nof object detection models. In Proceedings WACV. IEEE,\nJanuary 2005.\n[18] P. Roth and H. Bischof. On-line learning a person model from video\ndata. In Video Proc. CVPR, IEEE, 2006.\n[19] P.M. Roth, H. Grabner, D. Sko\u2566\u00e7caj, H. Bischof, and A. Leonardis. Online\nconservative learning for person detection. In Proceedings VS-PETS,\npages 223-230, Beijing, China, 2005. IEEE.\n[20] K. Sung and P. Niyogi. A formulation for active learning with\napplications to object detection. Technical Report AIM-1438, 1995.\n[21] A. Torralba, K.P. Murphy, and W.T. Freeman. Sharing features: Efficient\nboosting procedures for multiclass object detection. volume 2, pages\n762-769, 2005.\n[22] R. Verschae, J. Ruiz del Solar, and M. Correa. A unified learning\nframework for object detection and classification using nested cascades\nof boosted classifiers. Mach. Vision Appl., 19(2):85-103, 2008.\n[23] P. Viola and M. Jones. Rapid object detection using a boosted cascade\nof simple features. In CVPR, volume 1, pages 511-518, Kauai, HI,\nUSA, 2001. IEEE Computer Society.\n[24] Paul A. Viola and Michael J. Jones. Robust real-time face detection.\nIJCV, 57:137-154, 2004.\n[25] T. Yang, J. Li, Q. Pan, C. Zhao, and Y. Zhu. Active learning\nbased pedestrian detection in real scenes. In ICPR, pages 904-907,\nWashington, DC, USA, 2006. IEEE."]}

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.