Adaboost-Gaze Based Multi-Stage Object Tracking

Yu Bin Yang; Jiao Jiao Gu; Xiao Yu Zhang; Zhi Liu

doi:10.4028/www.scientific.net/AMM.246-247.179

Paper Titles

Weaken Polygon Effect and Improve Chain Drive Performance
p.159

Improved Genetic Optimization of Electric Motorcycle Planetary Transmission
p.164

Experimental Study of the Influence of Impact Mass on Damping Performance of Impact Dampers
p.169

A Frequency Measuring Method with Photoelectric Coupling Apparatus
p.175

Adaboost-Gaze Based Multi-Stage Object Tracking
p.179

MMW Radiometric Image Feature Extraction Method of Concealed Objects Based on Watershed Algorithm
p.184

Structure Analysis of FTIR Infrared Spectrometer Based on ANSYS/Workbench
p.189

Noninvasive Heart Rate Variability Detection Device for Fatigue Driving Detection System
p.194

A Real-Time Microseismic Monitoring System Based on Virtual Instruments
p.199

HomeApplied Mechanics and MaterialsApplied Mechanics and Materials Vols. 246-247Adaboost-Gaze Based Multi-Stage Object Tracking

Adaboost-Gaze Based Multi-Stage Object Tracking

Abstract:

Object tracking is a very important application in the fields of computer vision. In practice, automated tracking systems can rarely meet the required performance. This paper improves the attentional based tracking framework with Adaboost and gaze selection. The object classifier is implemented using the ADA Boosting to recognize digits from the MNIST dataset. At the initial object position Gaze selection is performed. The performance of the framework is evaluated using digit videos generated from the MNIST dataset with clutters. In general, the performance of the framework is robust to changes in motion routes and degree of clutter.

You might also be interested in these eBooks

Computer-Aided Design, Manufacturing, Modeling and Simulation II

View Preview

Info:

Periodical:

Applied Mechanics and Materials (Volumes 246-247)

Pages:

179-183

DOI:

https://doi.org/10.4028/www.scientific.net/AMM.246-247.179

Citation:

Cite this paper

Online since:

December 2012

Authors:

Yu Bin Yang, Jiao Jiao Gu, Xiao Yu Zhang, Zhi Liu

Keywords:

Adaboost, Computer Vision, Gaze, Kalman, MNIST, Object Tracking

Export:

RIS, BibTeX

Price:

Permissions CCC:

Request Permissions

Permissions PLS:

Request Permissions

Сopyright:

Citation:

References

[1] Rensink, Ronald A. The dynamic representation of scenes. Visual Cognition(2000), pp.17-42.

Google Scholar

[2] Bazzani L., Freitas N., Larochelle H. Learning Attentional Policies for Tracking and Recognition in Video with Deep Networks. Proceedings of the 28th International Conference on Machine Learning, (2011).

Google Scholar

[3] Auer P., Cesa-Bianchi N., Freund Y. Gambling in a rigged casino: the adversarial multi-armed andit problem. Technical Report NC2-TR-1998-025, (1998).

DOI: 10.1109/sfcs.1995.492488

Google Scholar

[4] H. Larochelle, G. Hinton. Learning to combine foveal glimpses with a third-order Boltzmann machine. In Neural Information Processing Systems, (2010).

Google Scholar

[5] Y. LeCun, L. Bottou, Y. Bengio, P. Haffner. Gradient-based learning applied to document recognition. Proceedings of the IEEE, Vol. 11(1998), PP. 2278-2324.

DOI: 10.1109/5.726791

Google Scholar

[6] Yoav Freund, Robert E. Schapire. A Decision-Theoretic Generalization of on-Line Learning and an Application to Boosting, (1995).

Google Scholar

[7] Yoav Freund, Corinna Cortes, http: /yann. lecun. com/exdb/mnist/, THE MNIST DATABASE of handwritten digits.

Google Scholar

[8] S. Belongie, M. Jitendra, J. Puzicha. Object Recognition Using Shape Contexts. IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 24, No. 24(2002).

DOI: 10.1109/34.993558

Google Scholar