Dissimilarity-based Classifiers Research Articles

SFINGE 3D: A novel benchmark for online detection and recognition of heterogeneous hand gestures from 3D fingers’ trajectories

In recent years gesture recognition has become an increasingly interesting topic for both research and industry. While interaction with a device through a gestural interface is a promising idea in several applications especially in the industrial field, some of the issues related to the task are still considered a challenge. In the scientific literature, a relevant amount of work has been recently presented on the problem of detecting and classifying gestures from 3D hands’ joints trajectories that can be captured by cheap devices installed on head-mounted displays and desktop computers. The methods proposed so far can achieve very good results on benchmarks requiring the offline supervised classification of segmented gestures of a particular kind but are not usually tested on the more realistic task of finding gestures execution within a continuous hand tracking session.In this paper, we present a novel benchmark, SFINGE 3D, aimed at evaluating online gesture detection and recognition. The dataset is composed of a dictionary of 13 segmented gestures used as a training set and 72 trajectories each containing 3–5 of the 13 gestures, performed in continuous tracking, padded with random hand movements acting as noise. The presented dataset, captured with a head-mounted Leap Motion device, is particularly suitable to evaluate gesture detection methods in a realistic use-case scenario, as it allows the analysis of online detection performance on heterogeneous gestures, characterized by static hand pose, global hand motions, and finger articulation.We exploited SFINGE 3D to compare two different approaches for the online detection and classification, one based on visual rendering and Convolutional Neural Networks and the other based on geometry-based handcrafted features and dissimilarity-based classifiers. We discuss the results, analyzing strengths and weaknesses of the methods, and deriving useful hints for their improvement.

Recognition of driver’s fatigue expression using Local Multiresolution Derivative Pattern

To develop the human-centric driver fatigue monitoring system for automatic understanding and charactering of driver's conditions, a novel, efficient feature extraction approach, named Local Multiresolution Derivative Pattern (LMDP), is proposed to describe the driver's fatigue expression images, and the Intersection Kernel Support Vector Machines classifier is then exploited to recognize three pre-defined classes of fatigue expressions, i.e., awake expressions, moderate fatigue expressions and severe fatigue expressions. With features extracted from a fatigue expressions dataset created at Southeast University, the holdout and cross-validation experiments on fatigue expressions classification are conducted by the Intersection Kernel Support Vector Machines classifier, compared with three commonly used classification methods including the k-nearest neighbor classifier, the multilayer perception classifier and the dissimilarity-based classifier. The experimental results of holdout and cross-validation showed that LMDP offers the better performance than Local Derivative Pattern, and the second order LMDP exceeds other order LMDP. With the second order LMDP and the Intersection Kernel Support Vector Machines classifier, the classification accuracies of the severe fatigue are over 90 in the holdout and cross-validation experiments, thus demonstrating the effectiveness of the proposed feature extraction method in automatically understanding the driver's conditions towards the human-centric driver fatigue monitoring system.

Mapping microarray gene expression data into dissimilarity spaces for tumor classification

Microarray gene expression data sets usually contain a large number of genes, but a small number of samples. In this article, we present a two-stage classification model by combining feature selection with the dissimilarity-based representation paradigm. In the preprocessing stage, the ReliefF algorithm is used to generate a subset with a number of top-ranked genes; in the learning/classification stage, the samples represented by the previously selected genes are mapped into a dissimilarity space, which is then used to construct a classifier capable of separating the classes more easily than a feature-based model. The ultimate aim of this paper is not to find the best subset of genes, but to analyze the performance of the dissimilarity-based models by means of a comprehensive collection of experiments for the classification of microarray gene expression data. To this end, we compare the classification results of an artificial neural network, a support vector machine and the Fisher’s linear discriminant classifier built on the feature (gene) space with those on the dissimilarity space when varying the number of genes selected by ReliefF, using eight different microarray databases. The results show that the dissimilarity-based classifiers systematically outperform the feature-based models. In addition, classification through the proposed representation appears to be more robust (i.e. less sensitive to the number of genes) than that with the conventional feature-based representation.

Recognizing Daily and Sports Activities in Two Open Source Machine Learning Environments Using Body-Worn Sensor Units

This study provides a comparative assessment on the different techniques of classifying human activities performed while wearing inertial and magnetic sensor units on the chest, arms and legs. The gyroscope, accelerometer and the magnetometer in each unit are tri-axial. Naive Bayesian classifier, artificial neural networks (ANNs), dissimilarity-based classifier, three types of decision trees, Gaussian mixture models (GMMs) and support vector machines (SVMs) are considered. A feature set extracted from the raw sensor data using principal component analysis is used for classification. Three different cross-validation techniques are employed to validate the classifiers. A performance comparison of the classifiers is provided in terms of their correct differentiation rates, confusion matrices and computational cost. The highest correct differentiation rates are achieved with ANNs (99.2%), SVMs (99.2%) and a GMM (99.1%). GMMs may be preferable because of their lower computational requirements. Regarding the position of sensor units on the body, those worn on the legs are the most informative. Comparing the different sensor modalities indicates that if only a single sensor type is used, the highest classification rates are achieved with magnetometers, followed by accelerometers and gyroscopes. The study also provides a comparison between two commonly used open source machine learning environments (WEKA and PRTools) in terms of their functionality, manageability, classifier performance and execution times.

A generalization of dissimilarity representations using feature lines and feature planes

Even though, under representational restrictions, the nearest feature rules and the dissimilarity-based classifiers are feasible alternatives to the nearest neighbor method; individually, they may not be sufficiently powerful if a very small set of prototypes is required, e.g. when it is computationally expensive to deal with larger sets of prototypes. In this paper, we show that combining both strategies, taking advantage of their individual properties, provides an improvement, particularly for correlated data sets. The combined strategy consists in deriving an enriched (generalized) dissimilarity representation by using the nearest feature rules, namely feature lines and feature planes. On top of that enriched representation, Bayesian classifiers can be constructed in order to obtain a good generalization.

On using prototype reduction schemes to optimize dissimilarity-based classification

The aim of this paper is to present a strategy by which a new philosophy for pattern classification, namely that pertaining to dissimilarity-based classifiers (DBCs), can be efficiently implemented. This methodology, proposed by Duin and his co-authors (see Refs. [Experiments with a featureless approach to pattern recognition, Pattern Recognition Lett. 18 (1997) 1159–1166; Relational discriminant analysis, Pattern Recognition Lett. 20 (1999) 1175–1181; Dissimilarity representations allow for buillding good classifiers, Pattern Recognition Lett. 23 (2002) 943–956; Dissimilarity representations in pattern recognition, Concepts, theory and applications, Ph.D. Thesis, Delft University of Technology, Delft, The Netherlands, 2005; Prototype selection for dissimilarity-based classifiers, Pattern Recognition 39 (2006) 189–208]), is a way of defining classifiers between the classes, and is not based on the feature measurements of the individual patterns, but rather on a suitable dissimilarity measure between them. The advantage of this methodology is that since it does not operate on the class-conditional distributions, the accuracy can exceed the Bayes’ error bound. The problem with this strategy is, however, the need to compute, store and process the inter-pattern dissimilarities for all the training samples, and thus, the accuracy of the classifier designed in the dissimilarity space is dependent on the methods used to achieve this. In this paper, we suggest a novel strategy to enhance the computation for all families of DBCs. Rather than compute, store and process the DBC based on the entire data set, we advocate that the training set be first reduced into a smaller representative subset. Also, rather than determine this subset on the basis of random selection, or clustering, etc., we advocate the use of a prototype reduction scheme (PRS), whose output yields the points to be utilized by the DBC. The rationale for this is explained in the paper. Apart from utilizing PRSs, in the paper we also propose simultaneously employing the Mahalanobis distance as the dissimilarity-measurement criterion to increase the DBCs classification accuracy. Our experimental results demonstrate that the proposed mechanism increases the classification accuracy when compared with the “conventional” approaches for samples involving real-life as well as artificial data sets— even though the resulting dissimilarity criterion is not symmetric.

Experimental study on prototype optimisation algorithms for prototype-based classification in vector spaces

Prototype-based classification relies on the distances between the examples to be classified and carefully chosen prototypes. A small set of prototypes is of interest to keep the computational complexity low, while maintaining high classification accuracy. An experimental study of some old and new prototype optimisation techniques is presented, in which the prototypes are either selected or generated from the given data. These condensing techniques are evaluated on real data, represented in vector spaces, by comparing their resulting reduction rates and classification performance. Usually the determination of prototypes is studied in relation with the nearest neighbour rule. We will show that the use of more general dissimilarity-based classifiers can be more beneficial. An important point in our study is that the adaptive condensing schemes here discussed allow the user to choose the number of prototypes freely according to the needs. If such techniques are combined with linear dissimilarity-based classifiers, they provide the best trade-off of small condensed sets and high classification accuracy.

Prototype selection for dissimilarity-based classifiers

A conventional way to discriminate between objects represented by dissimilarities is the nearest neighbor method. A more efficient and sometimes a more accurate solution is offered by other dissimilarity-based classifiers. They construct a decision rule based on the entire training set, but they need just a small set of prototypes, the so-called representation set, as a reference for classifying new objects. Such alternative approaches may be especially advantageous for non-Euclidean or even non-metric dissimilarities. The choice of a proper representation set for dissimilarity-based classifiers is not yet fully investigated. It appears that a random selection may work well. In this paper, a number of experiments has been conducted on various metric and non-metric dissimilarity representations and prototype selection methods. Several procedures, like traditional feature selection methods (here effectively searching for prototypes), mode seeking and linear programming are compared to the random selection. In general, we find out that systematic approaches lead to better results than the random selection, especially for a small number of prototypes. Although there is no single winner as it depends on data characteristics, the k-centres works well, in general. For two-class problems, an important observation is that our dissimilarity-based discrimination functions relying on significantly reduced prototype sets (3–10% of the training objects) offer a similar or much better classification accuracy than the best k-NN rule on the entire training set. This may be reached for multi-class data as well, however such problems are more difficult.