Classification with nonmetric distances: Image retrieval and class representation

被引:140
作者
Jacobs, DW
Weinshall, D
Gdalyahu, Y
机构
[1] NEC Res Inst, Princeton, NJ 08540 USA
[2] Hebrew Univ Jerusalem, Inst Comp Sci, IL-91904 Jerusalem, Israel
[3] IBM Res Corp, Haifa Lab, MATAM, Haifa, Israel
关键词
nonmetric; image retrieval; classification; supervised learning; median; condensing; nearest-neighbor; triangle inequality; robust distance; representation;
D O I
10.1109/34.862197
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
One of the key problems in appearance-based vision is understanding how to use a set of labeled images to classify new images. Classification systems that can model human performance, or that use robust image matching methods, often make use of similarity judgments that are nonmetric; but when the triangle inequality is not obeyed, most existing pattern recognition techniques are not applicable. We note that exemplar-based (or nearest-neighbor) methods can be applied naturally when using a wide class of nonmetric similarity functions. The key issue, however, is to find methods for choosing good representatives of a class that accurately characterize it. We show that existing condensing techniques for finding class representatives are ill-suited to deal with nonmetric dataspaces. We then focus on developing techniques for solving this problem, emphasizing two points: First, we show that the distance between two images is not a good measure of how well one image can represent another in nonmetric spaces. Instead, we use the vector correlation between the distances from each image to other previously seen images. Second, we show that in nonmetric spaces, boundary points are less significant for capturing the structure of a class than they are in Euclidean spaces. We suggest that atypical points may be more important in describing classes. We demonstrate the importance of these ideas to learning that generalizes from experience by improving performance using both synthetic and real images. In addition, we suggest ways of applying parametric techniques to supervised learning problems that involve a specific nonmetric distance functions, showing in particular how to generalize the idea of linear discriminant functions in a way that may be more useful in nonmetric spaces.
引用
收藏
页码:583 / 600
页数:18
相关论文
共 44 条
[1]  
[Anonymous], 1992, Entropy Optimization Principle with Applications
[2]   Determining the similarity of deformable shapes [J].
Basri, R ;
Costa, L ;
Geiger, D ;
Jacobs, D .
VISION RESEARCH, 1998, 38 (15-16) :2365-2385
[3]   The robust estimation of multiple motions: Parametric and piecewise-smooth flow fields [J].
Black, MJ ;
Anandan, P .
COMPUTER VISION AND IMAGE UNDERSTANDING, 1996, 63 (01) :75-104
[4]  
Blake A., 1987, Visual Reconstruction
[5]  
Blatt M, 1996, ADV NEUR IN, V8, P416
[6]  
BRAND M, 1996, 406 MIT MED LAB
[7]  
CORTES C, 1995, MACH LEARN, V20, P273, DOI 10.1023/A:1022627411411
[8]  
COX I, 1996, P INT C PATT REC, V100, P361
[9]   MINIMAL CONSISTENT SET (MCS) IDENTIFICATION FOR OPTIMAL NEAREST-NEIGHBOR DECISION SYSTEMS-DESIGN [J].
DASARATHY, BV .
IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS, 1994, 24 (03) :511-517
[10]   Sparse representations for image decomposition with occlusions [J].
Donahue, M ;
Geiger, D ;
Hummel, R ;
Liu, TL .
1996 IEEE COMPUTER SOCIETY CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, PROCEEDINGS, 1996, :7-12