1,720,966 research outputs found
Modelling frr of biometric verification systems using the template Co-update algorithm
No full textThe decrease of representativeness of available templates during time is due to the large intra-class variations characterizing biometrics (e.g. faces). This requires the design of algorithms able to make biometric verification systems adaptive to such variations. Among others, the template co-update algo-rithm, which uses the mutual help of two complementary biometric matchers, has shown promising experimental results. The present paper is aimed to describe a theoretical model able to explain the co-update behaviour. In particular, the focus is on the relationships between error rate and gallery size increase. Preliminary experimental results are shown to validate the proposed model. © Springer-Verlag Berlin Heidelberg 2009
Semi-supervised techniques for improving the performance of multiple classifier systems and personal recognition systems using biometric traits
No full textIn this paper, some research activities carried out by the Pattern Recognition and Applications Group of the University of Cagliari are presented. From the methodological viewpoint, the research activities of the group have been focused since its foundation on the field of pattern recognition. In this paper, we describe some research activities with particular reference to the development of semi-supervised algorithms applied to multiple classifier systems. From the viewpoint of applied research, the group is currently involved in the development of semi-supervised techniques for security applications. In this paper, we focus on the personal authentication/identification through biometrics. These research activities are currently funded by several contracts and grants from private firms and government agencies
Ensemble learning for Intrusion Detection in Computer Networks
No full textThe security of computer networks plays a strategic role in modern
computer systems. In order to enforce high protection levels against threats, a
number of software tools are currently developed. Intrusion Detection Systems aim
at detecting intruder who eluded the "first line" protection. In this paper, a pattern
recognition approach to network intrusion detection based on ensemble learning
paradigms is proposed. The potentialities of such an approach for data fusion and
some open issues are outline
Template co-update in multimodal biometric systems
No full textPerformances of biometric recognition systems can degrade quickly when the input biometric traits exhibit substantial variations compared to the templates collected during the enrolment stage of users. This issue can be addressed using template update methods. In this paper, a novel template update method based on the concept of biometric co-training is presented. In multimodal biometric systems, this method allows co-updating the template galleries of different biometrics, realizing a co-training process of biometric experts which allows updating templates more quickly and effectively. Reported results provide a first experimental evidence of the effectiveness of the proposed template update method. © Springer-Verlag Berlin Heidelberg 2007
On the variability of functional connectivity and network measures in source-reconstructed eeg time-series
Full text linkThe idea of estimating the statistical interdependence among (interacting) brain regions has motivated numerous researchers to investigate how the resulting connectivity patterns and networks may organize themselves under any conceivable scenario. Even though this idea has developed beyond its initial stages, its practical application is still far away from being widespread. One concurrent cause may be related to the proliferation of different approaches that aim to catch the underlying statistical interdependence among the (interacting) units. This issue has probably con-tributed to hindering comparisons among different studies. Not only do all these approaches go under the same name (functional connectivity), but they have often been tested and validated using different methods, therefore, making it difficult to understand to what extent they are similar or not. In this study, we aim to compare a set of different approaches commonly used to estimate the functional connectivity on a public EEG dataset representing a possible realistic scenario. As expected, our results show that source-level EEG connectivity estimates and the derived network measures, even though pointing to the same direction, may display substantial dependency on the (often ar-bitrary) choice of the selected connectivity metric and thresholding approach. In our opinion, the observed variability reflects the ambiguity and concern that should always be discussed when re-porting findings based on any connectivity metric
Eeg fingerprints under naturalistic viewing using a portable device
Full text linkThe electroencephalogram (EEG) has been proven to be a promising technique for personal identification and verification. Recently, the aperiodic component of the power spectrum was shown to outperform other commonly used EEG features. Beyond that, EEG characteristics may capture relevant features related to emotional states. In this work, we aim to understand if the aperiodic component of the power spectrum, as shown for resting-state experimental paradigms, is able to capture EEG-based subject-specific features in a naturalistic stimuli scenario. In order to answer this question, we performed an analysis using two freely available datasets containing EEG recordings from participants during viewing of film clips that aim to trigger different emotional states. Our study confirms that the aperiodic components of the power spectrum, as evaluated in terms of offset and exponent parameters, are able to detect subject-specific features extracted from the scalp EEG. In particular, our results show that the performance of the system was significantly higher for the film clip scenario if compared with resting-state, thus suggesting that under naturalistic stimuli it is even easier to identify a subject. As a consequence, we suggest a paradigm shift, from task-based or resting-state to naturalistic stimuli, when assessing the performance of EEG-based biometric systems
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
“Artificial histology” in colonic Neoplasia: A critical approach
Full text linkBackground: The histological assessment of colorectal precancer and cancer lesions is challenging and primarily impacts the clinical strategies of secondary colon cancer prevention. Artificial intelligence (AI) models may potentially assist in the histological diagnosis of this spectrum of phenotypical changes. Objectives: To provide a current overview of the evidence on AI-based methods for histologically assessing colonic precancer and cancer lesions. Methods: Based on the available studies, this review focuses on the reliability of AI-driven models in ranking the histological phenotypes included in colonic oncogenesis. Results: This review acknowledges the efforts to shift from subjective pathologists-based to more objective AI-based histological phenotyping. However, it also points out significant limitations and areas that require improvement. Conclusions: Current AI-driven methods have not yet achieved the expected level of clinical effectiveness, and there are still significant ethical concerns that need careful consideration. The integration of "artificial histology" into diagnostic practice requires further efforts to combine advancements in engineering techniques with the expertise of pathologists
- …
