The primary goal of pattern recognition is supervised or unsupervised classification. Among the various frameworks in which pattern recognition has been traditionally formulated, the statistical approach has been most intensively studied and used in practice. More recently, neural network techniques and methods imported from statistical learning theory have bean receiving increasing attention. The design of a recognition system requires careful attention to the following issues: definition of pattern classes, sensing environment, pattern representation, feature extraction and selection, cluster analysis, classifier design and learning, selection of training and test samples, and performance evaluation. In spite of almost 50 years of research and development in this field, the general problem of recognizing complex patterns with arbitrary orientation, location, and scale remains unsolved. New and emerging applications, such as data mining, web searching, retrieval of multimedia data, face recognition, and cursive handwriting recognition, require robust and efficient pattern recognition techniques. The objective of this review paper is to summarize and compare some of the well-known methods used in various stages of a pattern recognition system and identify research topics and applications which are at the forefront of this exciting and challenging field.

Abstract — We consider the problem of automatically recognizing human faces from frontal views with varying expression and illumination, as well as occlusion and disguise. We cast the recognition problem as one of classifying among multiple linear regression models, and argue that new theory from sparse signal representation offers the key to addressing this problem. Based on a sparse representation computed by ℓ 1-minimization, we propose a general classification algorithm for (image-based) object recognition. This new framework provides new insights into two crucial issues in face recognition: feature extraction and robustness to occlusion. For feature extraction, we show that if sparsity in the recognition problem is properly harnessed, the choice of features is no longer critical. What is critical, however, is whether the number of features is sufficiently large and whether the sparse representation is correctly computed. Unconventional features such as downsampled images and random projections perform just as well as conventional features such as Eigenfaces and Laplacianfaces, as long as the dimension of the feature space surpasses certain threshold, predicted by the theory of sparse representation. This framework can handle errors due to occlusion and corruption uniformly, by exploiting the fact that these errors are often sparse w.r.t. to the standard (pixel) basis. The theory of sparse representation helps predict how much occlusion the recognition algorithm can handle and how to choose the training images to maximize robustness to occlusion. We conduct extensive experiments on publicly available databases to verify the efficacy of the proposed algorithm, and corroborate the above claims.

Text which appears in a scene or is graphically added to video can provide an important supplemental source of index information as well as clues for decoding the video's structure and for classification. In this paper we present algorithms for detecting and tracking text in digital video. Our system implements a scalespace feature extractor that feeds an artificial neural processor to detect text blocks. Our text tracking scheme consists of two modules: an SSD (Sum of Squared Difference)-based module to find the initial position and a contour-based module to refine the position. Experiments conducted with a variety of video sources show that our scheme can detect and track text robustly.

Some previous works show that a web page can be partitioned to multiple segments or blocks, and usually the importance of those blocks in a page is not equivalent. Also, it is proved that differentiating noisy or unimportant blocks from pages can facilitate web mining, search and accessibility. But in these works, no uniform approach or model is presented to measure the importance of different portions in web pages. Through a user study, we found that people do have a consistent view about the importance of blocks in web pages. In this paper, we investigate how to find a model to automatically assign importance values to blocks in a web page. We define the block importance estimation as a learning problem. First, we use the VIPS (VIsion-based Page Segmentation) algorithm to partition a web page into semantic blocks with a hierarchy structure. Then spatial features (such as position, size) and content features (such as the number of images and links) are extracted to construct a feature vector for each block. Based on these features, learning algorithms, such as SVM and neural network, are applied to train various block importance models. In our experiments, the best model can achieve the performance with Micro-F1 79 % and Micro-Accuracy 85.9%, which is quite close to a person’s.

This paper surveys some recent theoretical results on the efficiency of machine learning algorithms. The main tool described is the notion of Probably Approximately Correct (PAC) learning, introduced by Valiant. We define this learning model and then look at some of the results obtained in it. We then consider some criticisms of the PAC model and the extensions proposed to address these criticisms. Finally, we look briefly at other models recently proposed in computational learning theory. 2 Introduction It's a dangerous thing to try to formalize an enterprise as complex and varied as machine learning so that it can be subjected to rigorous mathematical analysis. To be tractable, a formal model must be simple. Thus, inevitably, most people will feel that important aspects of the activity have been left out of the theory. Of course, they will be right. Therefore, it is not advisable to present a theory of machine learning as having reduced the entire field to its bare essentials. All ...

Spurred by developments in spatial planning in robotics, computer graphics, and VLSI layout, considerable attention has been devoted recently to the problem of moving sets of objects, such as line segments and polygons in the plane to polyhedra in three dimensions, without allowing collisions between the objects. One class of such problems considers the separability of sets of objects under different kinds of motions and various definitions of separation. This paper surveys this new area of research in a tutorial fashion, present new results, and provides a list of open problems and suggestions for further research. Key Words and Phrases: sofa problem, polygons, polyhedra, movable separability, visibility hulls, hidden lines, hidden surfaces, algorithms, complexity, computational geometry, spatial planning, collision avoidance, robotics, artificial intelligence. CR Categories: 3.36, 3.63, 5.25. 5.32. 5.5 * Research supported by NSERC Grant no. A9293 and FCAR Grant no.EQ1678. - 2 - ...

We present general sufficient conditions for the almost sure L1-consistency of his-togram density estimates based on data-dependent partitions. Analogous condi-tions guarantee the almost-sure risk consistency of histogram classification schemes based on data-dependent partitions. Multivariate data is considered throughout. In each case, the desired consistency requires shrinking cells, subexponential growth of a combinatorial complexity measure, and sub-linear growth of the num-ber of cells. It is not required that the cells of every partition be rectangles with sides paralles to the coordinate axis, or that each cell contain a minimum number of points. No assumptions are made concerning the common distribution of the training vectors. We apply the results to establish the consistency of several known partitioning estimates, including the kn-spacing density estimate, classifiers based on statisti-cally equivalent blocks, and classifiers based on multivariate clustering schemes.

Abstract not found

Two protocols used for learning under the pac-learning model introduced by Valiant are learning from random examples and learning from membership queries. Membership queries have also been used to efficiently and exactly learn a concept class that is too difficult to pac-learn using random examples. We ask whether using membership queries-- in conjunction with or instead of random examples serve a new purpose: helping to reduce the total number of examples needed to pac-learn a concept class C already known to be pac-learnable using just random examples. We focus on concept classes that are dense in themselves, such as haft-spaces of R ', rectangles in the plane, and the class Z = {[0, a]: 0 _ a < 1} of initial segments of [0, 1]. The main

We introduce a novel optimization method based on semidefinite programming relaxations to the field of computer vision and apply it to the combinatorial problem of minimizing quadratic functionals in binary decision variables subject to linear constraints.