Learning to Recognize 3d Objects in Presence of Uncertainty

The subject of this paper is object recognition, for the case in which uncertain parametric representations of objects are available. The use of learning via connectionist architectures is investigated in order to solve the problem of mapping recovered shape parameters into a probability measure over a set of known objects, taking explicitly into account estimates of uncertainty in the parameter themselves. The additional insight thus gained on the properties of the recovered shape data, from the point of view of the recognition task, qualiies this method as an independent, performance oriented, validation tool for complex shape recovery systems. The feasibility of the approach is veriied through extensive simulation experiments. Promising results thus obtained are reported.

[1]  Elijah Polak,et al.  Computational methods in optimization , 1971 .

[2]  Barr,et al.  Superquadrics and Angle-Preserving Transformations , 1981, IEEE Computer Graphics and Applications.

[3]  David B. Cooper,et al.  On Optimally Combining Pieces of Information, with Application to Estimating 3-D Complex-Object Position from Range Data , 1986, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[4]  John Scott Bridle,et al.  Probabilistic Interpretation of Feedforward Classification Network Outputs, with Relationships to Statistical Pattern Recognition , 1989, NATO Neurocomputing.

[5]  Naftali Tishby,et al.  Consistent inference of probabilities in layered networks: predictions and generalizations , 1989, International 1989 Joint Conference on Neural Networks.

[6]  Kurt Hornik,et al.  Multilayer feedforward networks are universal approximators , 1989, Neural Networks.

[7]  Ruzena Bajcsy,et al.  Recovery of Parametric Models from Range Images: The Case for Superquadrics with Global Deformations , 1990, IEEE Trans. Pattern Anal. Mach. Intell..

[8]  Edoardo Ardizzone,et al.  An Associative Link from Geometric to Symbolic Representations in Artificial Vision , 1991, AI*IA.

[9]  David J. C. MacKay,et al.  The Evidence Framework Applied to Classification Networks , 1992, Neural Computation.

[10]  David J. C. MacKay,et al.  Bayesian Interpolation , 1992, Neural Computation.

[11]  Elie Bienenstock,et al.  Neural Networks and the Bias/Variance Dilemma , 1992, Neural Computation.

[12]  Edoardo Ardizzone,et al.  The Mapping Between Geometric and Symbolic Knowledge in Artificial Vision: A Connectionist Approach , 1992, ECAI.

[13]  Anil K. Jain,et al.  Recognizing geons from superquadrics fitted to range data , 1992, Image Vis. Comput..

[14]  Martin Fodslette Møller,et al.  A scaled conjugate gradient algorithm for fast supervised learning , 1993, Neural Networks.

[15]  Frank P. Ferrie,et al.  Darboux Frames, Snakes, and Super-Quadrics: Geometry from the Bottom Up , 1993, IEEE Trans. Pattern Anal. Mach. Intell..

[16]  Frank P. Ferrie,et al.  Recognizing volumetric objects in the presence of uncertainty , 1994, Proceedings of 12th International Conference on Pattern Recognition.

[17]  Frank P. Ferrie,et al.  Autonomous exploration: driven by uncertainty , 1994, 1994 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.

[18]  David B. Cooper,et al.  Practical Reliable Bayesian Recognition of 2D and 3D Objects Using Implicit Polynomials and Algebraic Invariants , 1996, IEEE Trans. Pattern Anal. Mach. Intell..

[19]  Duan Li,et al.  On Restart Procedures for the Conjugate Gradient Method , 2004, Numerical Algorithms.