A Method for Lifelong Gesture Learning Based on Growing Neural Gas

Gesture-based interfaces offer the possibility of an intuitive command language for assistive robotics and ubiquitous computing. As an individual’s health changes with age, their ability to consistently perform standard gestures may decrease, particularly towards the end of life. Thus, such interfaces will need to be capable of learning commands which are not choreographed ahead of time by the system designers. This circumstance illustrates the need for a system which engages in lifelong learning and is capable of discerning new gestures and the user’s desired response to them. This paper describes an innovative approach to lifelong learning based on clustered gesture representations identified through the Growing Neural Gas algorithm. The simulated approach utilizes a user-generated reward signal to progressively refine the response of an assistive robot toward a preferred goal configuration.

[1]  Teuvo Kohonen,et al.  The self-organizing map , 1990 .

[2]  Thomas Villmann,et al.  Vector Quantization by Optimal Neural Gas , 1997, ICANN.

[3]  J. L. Hemmen,et al.  Nonlinear neural networks. , 1986, Physical review letters.

[4]  Shen Furao,et al.  An incremental network for on-line unsupervised classification and topology learning , 2006, Neural Networks.

[5]  Stephen Grossberg,et al.  Nonlinear neural networks: Principles, mechanisms, and architectures , 1988, Neural Networks.

[6]  Bernd Fritzke,et al.  A Growing Neural Gas Network Learns Topologies , 1994, NIPS.

[7]  Mubarak Shah,et al.  View-Invariant Representation and Recognition of Actions , 2002, International Journal of Computer Vision.

[8]  Bernd Fritzke,et al.  A Self-Organizing Network that Can Follow Non-stationary Distributions , 1997, ICANN.

[9]  Paul Yanik,et al.  Gesture-Based Robot Path Shaping , 2013 .

[10]  Ian D. Walker,et al.  A Gesture Learning Interface for Simulated Robot Path Shaping With a Human Teacher , 2014, IEEE Transactions on Human-Machine Systems.

[11]  Ernesto Estrada,et al.  The Structure of Complex Networks: Theory and Applications , 2011 .

[12]  Fred Henrik Hamker,et al.  Life-long learning Cell Structures--continuously learning without catastrophic interference , 2001, Neural Networks.

[13]  Claude F. Touzet,et al.  Neural reinforcement learning for behaviour synthesis , 1997, Robotics Auton. Syst..

[14]  Pierre-Yves Oudeyer,et al.  Robotic clicker training , 2002, Robotics Auton. Syst..