The elements of multimodal authentication along with system models are presented. These include the machine experts as well as machine supervisors. In particular fingerprint and speech based systems will serve as illustration of a mobile authentication application. A novel signal adaptive supervisor, based on the input biometric signal quality is evaluated. Experimental results on data collected from mobile telephones are reported demonstrating the benefits of the proposed scheme .

An on-line signature verification system exploiting both local and global information through decision-level fusion is presented. Global

We investigate four previously unexplored aspects of ensemble selection, a procedure for building ensembles of classifiers. First we test whether adjusting model predictions to put them on a canonical scale makes the ensembles more effective. Second, we explore the performance of ensemble selection when different amounts of data are available for ensemble hillclimbing. Third, we quantify the benefit of ensemble selection’s ability to optimize to arbitrary metrics. Fourth, we study the performance impact of pruning the number of models available for ensemble selection. Based on our results we present improved ensemble selection methods that double the benefit of the original method. 1

Databases are a key technology for molecular biology which is a very data intensive discipline. Since molecular biological databases are rather heterogeneous, unification and data integration is mandatory to make use of the huge amount of available information. Currently, the most promising approach for integration is the use of ontologies. Since mapping biological entities into ontologies is usually achieved manually or semiautomatically, a system for automatic classification of biological entities into ontologies saves time and effort. Therefore, we present a support vector machine based approach that automatically classifies biological entities into a given ontology. To solve this difficult task, our method copes with the following aspects. Biological entities might belong to more than one class or may be placed in classes on varying abstraction levels. An object may be described by several representations. Thus, the classifier has to be enabled to draw information from all of them, but must consider the possibility that some objects are described incompletely. Therefore, our method introduces the technique of objectadjusted weighting which regulates the impact of each representation dynamically for each object. To significantly improve the time performance of the classifier we exploit the inheritance relations of the given ontology. Our experimental evaluation on protein data and several parts of an established molecular biological ontology shows that our prototype offers impressive accuracy and is efficient enough to cope with the large number of classes encountered in real world problems. ∗ Supported by the German Ministery for Education, Science,

Classification methods based on learning from examples have been widely applied to character recognition from the 1990s and have brought forth significant improvements of recognition accuracies. This class of methods includes statistical methods, artificial neural networks, support vector machines, multiple classifier combination, etc. In this paper, we discuss the characteristics of the classification methods that have been successfully applied to character recognition, and show the remaining problems that can be potentially solved by learning methods. 1.

Different strategies to learn user semantic queries from dissimilarity representations of video audio-visual content are presented. When dealing with large corpora of videos documents, using a feature representation requires the online computation of distances between all documents and a query. Hence, a dissimilarity representation may be preferred because its offline computation speeds up the retrieval process. We show how distances related to visual and audio video features can directly be used to learn complex concepts from a set of positive and negative examples provided by the user. Based on the idea of dissimilarity spaces, we derive three algorithms to fuse modalities and therefore to enhance the precision of retrieval results. The evaluation of our technique is performed on artificial data and on the complete annotated TRECVID corpus.

Conventional approaches to combining classifiers improve accuracy at the cost of increased processing. We propose a novel search based approach to automatically combine multiple classifiers in a cascade to obtain the desired tradeoff between classification speed and classification accuracy. The search procedure only updates the rejection thresholds (one for each constituent classier) in the cascade, consequently no new classifiers are added and no training is necessary. A branch-and-bound version of depth-first-search with efficient pruning is proposed for finding the optimal thresholds for the cascade. It produces optimal solutions under arbitrary user specified speed and accuracy constraints. The effectiveness of the approach is demonstrated on handwritten character recognition by finding a) the fastest possible combination given an upper bound on classification error, and also b) the most accurate combination given a lower bound on speed.

In this work, we describe a speaker identification system that uses multiple supplementary information sources for computing a combined match score for the unknown speaker. Each speaker profile in the database consists of multiple feature vector sets that can vary in their scale, dimensionality, and the number of vectors. The evidence from a given feature set is weighted by its reliability that is set in a priori fashion. The confidence of the identification result is also estimated. The system is evaluated with a corpus of 110 Finnish speakers. The evaluated feature sets include mel-cepstrum, LPC-cepstrum, dynamic cepstrum, long-term averaged spectrum of /A/ vowel, and F0.

Statistical pattern recognition traditionally relies on a feature representation. This approach can be powerful, if sufficient knowledge is available to select a small set of well-discriminating features. If there is a lack of such knowledge, a large set of possible features has to be collected and a large training set, representative in distribution for the given problem, is needed to build a reliable classifier. This is partially caused by the inherent difficulties in the feature based representation, when a (large) set of suboptimal features is used, in may result in a class overlap and strong feature dependency. The dissimilarity representation aims at treating objects in their wholeness, avoiding the use of isolated features. If the dissimilarity measure is defined such that a zero value is only permitted for identical objects, class overlap may be avoided. Consequently, proper knowledge of class densities is not needed, which opens the possibility to a domain based classification in which the training set should be just representative for the domain of the classes. In this paper, first, the basic ideas and some results of the dissimilarity representation are summarized. It is followed by a discussion on how this may be worked out for the domain based pattern recognition. 1. Issues of pattern recognition In general, pattern recognition relies on the description of regularities in observations of classes of objects.

In this paper we present a combining strategy to cope with the problem of classification in ill-defined domains. In these cases, even though a particular target class may be sampled in a representative manner, an outlier class may be poorly sampled, or new outlier classes may occur that have not been considered during training. This may have a considerable impact on classification performance. The objective of a classifier in this situation is to utilise all known information in discriminating, and to remain as robust as possible to changing conditions. A classification scheme is presented that deals with this problem, consisting of a sequential combination of a one-class and multi-class classifier. We show that it can outperform the traditional classifier with reject-option scheme, locally selecting/training models for the purpose of optimising the classification and rejection performance.