Predicting Performance of Object Recognition

We present a method for predicting fundamental performance of object recognition. We assume that both scene data and model objects are represented by 2D point features and a data/model match is evaluated using a vote-based criterion. The proposed method considers data distortion factors such as uncertainty, occlusion, and clutter, in addition to model similarity. This is unlike previous approaches, which consider only a subset of these factors. Performance is predicted in two stages. In the first stage, the similarity between every pair of model objects is captured by comparing their structures as a function of the relative transformation between them. In the second stage, the similarity information is used along with statistical models of the data-distortion factors to determine an upper bound on the probability of recognition error. This bound is directly used to determine a lower bound on the probability of correct recognition. The validity of the method is experimentally demonstrated using real synthetic aperture radar (SAR) data.

[1]  Anil K. Jain,et al.  Integration of Multiple Feature Groups and Multiple Views into a 3D Object Recognition System , 1995, Comput. Vis. Image Underst..

[2]  David G. Lowe,et al.  Learning Appearance Models for Object Recognition , 1996, Object Representation in Computer Vision.

[3]  Karen B. Sarachik.  The Effect of Gaussian Error in Object Recognition , 1997, IEEE Trans. Pattern Anal. Mach. Intell..

[4]  M. Hebert,et al.  The Representation, Recognition, and Locating of 3-D Objects , 1986 .

[5]  George C. Stockman,et al.  Object recognition and localization via pose clustering , 1987, Comput. Vis. Graph. Image Process..

[6]  Michel Dhome,et al.  Determination of the Attitude of 3D Objects from a Single Perspective View , 1989, IEEE Trans. Pattern Anal. Mach. Intell..

[7]  Bir Bhanu,et al.  Bounding SAR ATR performance based on model similarity , 1999, Defense, Security, and Sensing.

[8]  W. Eric L. Grimson,et al.  Localizing Overlapping Parts by Searching the Interpretation Tree , 1987, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[9]  W. Eric L. Grimson,et al.  Verifying model-based alignments in the presence of uncertainty , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[10]  Bir Bhanu,et al.  Recognizing MSTAR target variants and articulations , 1999, Defense, Security, and Sensing.

[11]  W. Eric L. Grimson,et al.  On the Verification of Hypothesized Matches in Model-Based Recognition , 1991, IEEE Trans. Pattern Anal. Mach. Intell..

[12]  Michael Lindenbaum.  Bounds on Shape Recognition Performance , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[13]  Daniel P. Huttenlocher,et al.  A new Bayesian framework for object recognition , 1999, Proceedings. 1999 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No PR00149).

[14]  Anil K. Jain,et al.  Integration of Multiple Feature Groups and Multiple Views into a 3D Object Recognition System , 1993, Applications of Invariance in Computer Vision.

[15]  Thomas M. Breuel,et al.  Higher-order statistics in object recognition , 1993, Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.

[16]  Stan Z. Li,et al.  Markov Random Field Modeling in Computer Vision , 1995, Computer Science Workbench.

[17]  Bir Bhanu,et al.  Recognition of Articulated and Occluded Objects , 1999, IEEE Trans. Pattern Anal. Mach. Intell..

[18]  Michael Lindenbaum,et al.  An Integrated Model for Evaluating the Amount of Data Required for Reliable Recognition , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[19]  Michael Lindenbaum,et al.  Grouping-Based Nonadditive Verification , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[20]  Michael Lee Bryant,et al.  Standard SAR ATR evaluation experiments using the MSTAR public release data set , 1998, Defense, Security, and Sensing.