A Hierarchical RBF Online Learning Algorithm for Real-Time 3-D Scanner

In this paper, a novel real-time online network model is presented. It is derived from the hierarchical radial basis function (HRBF) model and it grows by automatically adding units at smaller scales, where the surface details are located, while data points are being collected. Real-time operation is achieved by exploiting the quasi-local nature of the Gaussian units: through the definition of a quad-tree structure to support their receptive field local network reconfiguration can be obtained. The model has been applied to 3-D scanning, where an updated real-time display of the manifold to the operator is fundamental to drive the acquisition procedure itself. Quantitative results are reported, which show that the accuracy achieved is comparable to that of two batch approaches: batch HRBF and support vector machines (SVMs). However, these two approaches are not suitable to real-time online learning. Moreover, proof of convergence is also given.

[1]  M. Hasenjäger,et al.  Active learning in neural networks , 2002 .

[2]  Peter Craven,et al.  Smoothing noisy data with spline functions , 1978 .

[3]  F. Girosi,et al.  Networks for approximation and learning , 1990, Proc. IEEE.

[4]  Noboru Murata,et al.  A Statistical Study on On-line Learning , 1999 .

[5]  John Moody,et al.  Fast Learning in Networks of Locally-Tuned Processing Units , 1989, Neural Computation.

[6]  Marc Levoy,et al.  QSplat: a multiresolution point rendering system for large meshes , 2000, SIGGRAPH.

[7]  Babak Hassibi,et al.  Second Order Derivatives for Network Pruning: Optimal Brain Surgeon , 1992, NIPS.

[8]  D. Saad,et al.  Dynamics of on-line learning in radial basis function networks , 1997 .

[9]  Haralambos Sarimveis,et al.  A new algorithm for online structure and parameter adaptation of RBF networks , 2003, Neural Networks.

[10]  Bernd Fritzke,et al.  Growing cell structures--A self-organizing network for unsupervised and supervised learning , 1994, Neural Networks.

[11]  J.C. Chedjou,et al.  Comparison of gradient descent method, Kalman filtering and decoupled Kalman in training neural networks used for fingerprint-based positioning , 2004, IEEE 60th Vehicular Technology Conference, 2004. VTC2004-Fall. 2004.

[12]  Robert M. Sanner,et al.  Gaussian Networks for Direct Adaptive Control , 1991, 1991 American Control Conference.

[13]  LinChih-Jen,et al.  A tutorial on -support vector machines , 2005 .

[14]  N. Alberto Borghese,et al.  A portable modular system for automatic acquisition of 3D objects , 2000, IEEE Trans. Instrum. Meas..

[15]  Lakhmi C. Jain,et al.  New Learning Paradigms in Soft Computing , 2002 .

[16]  Nello Cristianini,et al.  An Introduction to Support Vector Machines and Other Kernel-based Learning Methods , 2000 .

[17]  Heekuck Oh,et al.  Neural Networks for Pattern Recognition , 1993, Adv. Comput..

[18]  Vincenzo Piuri,et al.  Multiscale models for data processing: an experimental sensitivity analysis , 2001, IEEE Trans. Instrum. Meas..

[19]  James Theiler,et al.  Accurate On-line Support Vector Regression , 2003, Neural Computation.

[20]  Tomaso A. Poggio,et al.  Regularization Theory and Neural Networks Architectures , 1995, Neural Computation.

[21]  Bernhard Schölkopf,et al.  A tutorial on v-support vector machines , 2005 .

[22]  David Saad,et al.  Online Learning in Radial Basis Function Networks , 1997, Neural Computation.

[23]  Vincenzo Piuri,et al.  Automatic multiscale meshing through HRBF networks , 2005, IEEE Transactions on Instrumentation and Measurement.

[24]  Anton Gunzinger,et al.  Fast neural net simulation with a DSP processor array , 1995, IEEE Trans. Neural Networks.

[25]  Chih-Jen Lin,et al.  A tutorial on?-support vector machines , 2005 .

[26]  N. Alberto Borghese,et al.  Hierarchical RBF networks and local parameters estimate , 1998, Neurocomputing.

[27]  Y Lu,et al.  A Sequential Learning Scheme for Function Approximation Using Minimal Radial Basis Function Neural Networks , 1997, Neural Computation.

[28]  John Hallam,et al.  Combining Regression Trees and Radial Basis Function Networks , 2000, Int. J. Neural Syst..

[29]  David Saad,et al.  On-Line Learning in Neural Networks , 1999 .

[30]  Ahmed Rubaai,et al.  Online training of parallel neural network estimators for control of induction motors , 2001 .

[31]  Alexander J. Smola,et al.  Support Vector Regression Machines , 1996, NIPS.

[32]  Mark H. Lee,et al.  Error-driven active learning in growing radial basis function networks for early robot learning , 2006, Proceedings 2006 IEEE International Conference on Robotics and Automation, 2006. ICRA 2006..

[33]  Jun Fan,et al.  Online face recognition system for videos based on modified probabilistic neural networks , 2004, 2004 International Conference on Image Processing, 2004. ICIP '04..

[34]  Christopher M. Bishop,et al.  Neural networks for pattern recognition , 1995 .

[35]  Giancarlo Ferrigno,et al.  Autoscan: A Flexible and Portable 3D Scanner , 1998, IEEE Computer Graphics and Applications.

[36]  John C. Platt.  A Resource-Allocating Network for Function Interpolation , 1991, Neural Computation.

[37]  Mauro Maggioni,et al.  Multiscale approximation with hierarchical radial basis functions networks , 2004, IEEE Transactions on Neural Networks.

[38]  Giancarlo Ferrigno,et al.  Reducing and Filtering Point Clouds With Enhanced Vector Quantization , 2007, IEEE Transactions on Neural Networks.

[39]  Ganesh K. Venayagamoorthy,et al.  Online Training of a Generalized Neuron with Particle Swarm Optimization , 2006, The 2006 IEEE International Joint Conference on Neural Network Proceedings.

[40]  Marc Levoy,et al.  Real-time 3D model acquisition , 2002, ACM Trans. Graph..

[41]  Giovanni Montana,et al.  Learning to Trade with Incremental Support Vector Regression Experts , 2008, HAIS.

[42]  Thorsten Joachims,et al.  Making large scale SVM learning practical , 1998 .

[43]  J.I. Mulero-Martinez,et al.  Best Approximation of Gaussian Neural Networks With Nodes Uniformly Spaced , 2008, IEEE Transactions on Neural Networks.