An Experimental Platform Based on Transfer Learning Paradigm for Mobile Robots

Abstract:

The transfer learning paradigm promises to improve the adaptability of mobile robots which apply machine learning (ML) algorithms for perception and localization. An system framework based on transfer learning paradigm for mobile robots was proposed. The keys of the framework are 1) how to arrange the knowledge database, 2) how to select the source data, source model and priori knowledge, 3) the method to achieve the target model and 4) the method to gain new samples. An experimental platform was built up in order to gain data to verify the framework, whose underpan is the same as the Freescale Cup, MCUs core is ARM Cortex-M4, raw data is stored in a SD card and model is trained offline by a more powerful computer using ML algorithms. The MCU communicates with the powerful computer by Zigbee or WiFi. The experimental platform is cheap and feasible, which is validated by data collection and model training experiments.

• Periodical: Applied Mechanics and Materials (Volumes 336-338)
• Pages: 960-963
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.336-338.960
• Online since: July 2013
• Authors: Hong Bo Lv, Jun Jie Tu, Qi Guo Sun, Fang Yuan Li, Jiao Wang, Yan Xiao Lei
• Keywords: Experimental Platform, Mobile Robot, Transfer Learning

References

[1] F. Bonin-Font, A. Ortiz, and G. Oliver: Journal of Intelligent & Robotic Systems, vol. 53, no. 3, pp.263-296, (2008).

[2] A. Bar Hillel, R. Lerner, D. Levi, and G. Raz: Machine Vision and Applications, 2012 (online only, to be published).

[3] V. N. Vapnik: Statistical Learning Theory (John Wiley & Sons, Inc., 1998).

[4] S. J. Pan and Q. Yang: IEEE Transactions On Knowledge And Data Engineering, vol. 22, no. 10, pp.1345-1359, (2010).

[5] M. -R. Amini, A. Habrard, L. Ralaivola, and N. Usunier: Learning from non IID Data: Theory, Algorithms and Practice, in ECML PKDD 2009, Bled, Slovenia, (2009).

[6] M. E. Taylor and P. Stone: Journal of Machine Learning Research, vol. 10, pp.1633-1685, (2009).

[7] F. Lauer and G. Bloch: Neurocomputing, vol. 71, no. 7-9, pp.1578-1594, (2008).

[8] F. Lauer and G. Bloch: Machine Learning, vol. 70, no. 1, pp.89-118, (2008).

[9] B. Settles: Active learning literature survey (University of Wisconsin, Madison, 2010).

[10] B. D. Argall, S. Chernova, M. Veloso, and B. Browning: Robotics and Autonomous Systems, vol. 57, no. 5, pp.469-483, (2009).

DOI: 10.1016/j.robot.2008.10.024

[11] K. P. Murphy: Machine Learning: a Probabilistic Perspective (The MIT Press, 2012).

[12] V. N. Vapnik: The Nature of Statistical Learning Theory, 2nd ed. (Springer, 2000).