Vision-based representation and recognition of human activities in image sequences

Magdeburg, Univ., Fak. für Elektrotechnik und Informationstechnik, Diss., 2013

Rotation Invariant Real-time Face Detection and Recognition System

In this report, a face recognition system that is capable of detecting and recognizing frontal and rotated faces was developed. Two face recognition methods focusing on the aspect of pose invariance are presented and evaluated - the whole face approach and the component-based approach. The main challenge of this project is to develop a system that is able to identify faces under different viewing angles in realtime. The development of such a system will enhance the capability and robustness of current face recognition technology. The whole-face approach recognizes faces by classifying a single feature vector consisting of the gray values of the whole face image. The component-based approach first locates the facial components and extracts them. These components are normalized and combined into a single feature vector for classification. The Support Vector Machine (SVM) is used as the classifier for both approaches. Extensive tests with respect to the robustness against pose changes are performed on a database that includes faces rotated up to about 40 degrees in depth. The component-based approach clearly outperforms the whole-face approach on all tests. Although this approach isproven to be more reliable, it is still too slow for real-time applications. That is the reason why a real-time face recognition system using the whole-face approach is implemented to recognize people in color video sequences.

Face encoding and recognition in the human brain.

A dissociation between human neural systems that participate in the encoding and later recognition of new memories for faces was demonstrated by measuring memory task-related changes in regional cerebral blood flow with positron emission tomography. There was almost no overlap between the brain structures associated with these memory functions. A region in the right hippocampus and adjacent cortex was activated during memory encoding but not during recognition. The most striking finding in neocortex was the lateralization of prefrontal participation. Encoding activated left prefrontal cortex, whereas recognition activated right prefrontal cortex. These results indicate that the hippocampus and adjacent cortex participate in memory function primarily at the time of new memory encoding. Moreover, face recognition is not mediated simply by recapitulation of operations performed at the time of encoding but, rather, involves anatomically dissociable operations.

An incremental multilinear system for human face learning and recognition

This dissertation establishes a novel system for human face learning and recognition based on incremental multilinear Principal Component Analysis (PCA). Most of the existing face recognition systems need training data during the learning process. The system as proposed in this dissertation utilizes an unsupervised or weakly supervised learning approach, in which the learning phase requires a minimal amount of training data. It also overcomes the inability of traditional systems to adapt to the testing phase as the decision process for the newly acquired images continues to rely on that same old training data set. Consequently when a new training set is to be used, the traditional approach will require that the entire eigensystem will have to be generated again. However, as a means to speed up this computational process, the proposed method uses the eigensystem generated from the old training set together with the new images to generate more effectively the new eigensystem in a so-called incremental learning process. In the empirical evaluation phase, there are two key factors that are essential in evaluating the performance of the proposed method: (1) recognition accuracy and (2) computational complexity. In order to establish the most suitable algorithm for this research, a comparative analysis of the best performing methods has been carried out first. The results of the comparative analysis advocated for the initial utilization of the multilinear PCA in our research. As for the consideration of the issue of computational complexity for the subspace update procedure, a novel incremental algorithm, which combines the traditional sequential Karhunen-Loeve (SKL) algorithm with the newly developed incremental modified fast PCA algorithm, was established. In order to utilize the multilinear PCA in the incremental process, a new unfolding method was developed to affix the newly added data at the end of the previous data. The results of the incremental process based on these two methods were obtained to bear out these new theoretical improvements. Some object tracking results using video images are also provided as another challenging task to prove the soundness of this incremental multilinear learning method.

Representation and learning of visual information for pose recognition

Recovering position from sensor information is an important problem in mobile robotics, known as localisation. Localisation requires a map or some other description of the environment to provide the robot with a context to interpret sensor data. The mobile robot system under discussion is using an artificial neural representation of position. Building a geometrical map of the environment with a single camera and artificial neural networks is difficult. Instead it would be simpler to learn position as a function of the visual input. Usually when learning images, an intermediate representation is employed. An appropriate starting point for biologically plausible image representation is the complex cells of the visual cortex, which have invariance properties that appear useful for localisation. The effectiveness for localisation of two different complex cell models are evaluated. Finally the ability of a simple neural network with single shot learning to recognise these representations and localise a robot is examined.

Normal ECG Recognition for Express-Diagnostics Based on Scale-Space Representation and Dynamic Matching

A novel approach of normal ECG recognition based on scale-space signal representation is proposed. The approach utilizes curvature scale-space signal representation used to match visual objects shapes previously and dynamic programming algorithm for matching CSS representations of ECG signals. Extraction and matching processes are fast and experimental results show that the approach is quite robust for preliminary normal ECG recognition.

Effects of temporal association on recognition memory

The influence of temporal association on the representation and recognition of objects was investigated. Observers were shown sequences of novel faces in which the identity of the face changed as the head rotated. As a result, observers showed a tendency to treat the views as if they were of the same person. Additional experiments revealed that this was only true if the training sequences depicted head rotations rather than jumbled views; in other words, the sequence had to be spatially as well as temporally smooth. Results suggest that we are continuously associating views of objects to support later recognition, and that we do so not only on the basis of the physical similarity, but also the correlated appearance in time of the objects.

Effects of temporal association on recognition memory

The influence of temporal association on the representation and recognition of objects was investigated. Observers were shown sequences of novel faces in which the identity of the face changed as the head rotated. As a result, observers showed a tendency to treat the views as if they were of the same person. Additional experiments revealed that this was only true if the training sequences depicted head rotations rather than jumbled views: in other words, the sequence had to be spatially as well as temporally smooth. Results suggest that we are continuously associating views of objects to support later recognition, and that we do so not only on the basis of the physical similarity, but also the correlated appearance in time of the objects.

Characterization of the immunological properties of " Plasmodium falciparum " and " Plasmodium berghei " circumsporozoite proteins : antibody responses and recognition by specific CD8+ T cells

SUMMARY Interest in developing intervention strategies against malaria by targeting the liver stage of the Plasmodium life cycle has been fueled by studies which show that sterile protective immunity can be achieved by immunization with radiation-attenuated sporozoites. Anti-malarial drugs and insecticides have been widely used to control the disease, but in the hope of developing a more cost-effective intervention strategy, vaccine development has taken centre stage in malaria research. There is currently no vaccine against malaria. Attenuated sporozoite-induced immunity is achieved by antibodies and T cells against malaria liver stage antigens, the most abundant being the circumsporozoite protein (CSP), and many vaccine formulations aim at mimicking this immunity. However, the mechanisms by which the antibody and T cell immune responses are generated after infection by sporozoites, or after immunization with different vaccine formulations are still not well understood. The first part of this work aimed at determining the ability of primary hepatocytes from BALB/c mice to process and present CSP-derived peptides after infection with P. berghei sporozoites. Both infected hepatocytes and those traversed by sporozoites during migration were found to be capable of processing and presenting the CSP to specific CD8+ T cells in vitro. The pathway of processing and presentation involved the proteasome, aspartic proteases and transport through a post-Endoplasmic Reticulum (ER) compartment. These results suggest that in vivo, infected hepatocytes contribute to the elicitation and expansion of a T cell response. In the second part, the antibody responses of CB6F1 mice to synthetic peptides corresponding to the N- and C-terminal domains of P. berghei and P. falciparum CS proteins were characterized. Mice were immunized with single peptides or a combination of N- and C-terminal peptides. The peptides were immunogenic in mice and the antisera generated could recognize the native CSP on the sporozoite surface. Antisera generated against the N-terminal peptides or against the combinations inhibited sporozoite invasion of hepatocytes in vitro. In vivo, more mice immunized with single P. berghei peptides were protected from infection upon a challenge with P. berghei sporozoites, than mice immunized with a combination of N- and C-terminal peptides. Furthermore, P. falciparum N-terminal peptides were recognized by serum samples from people living in malaria-endemic areas. Importantly, recognition of a peptide from the N-terminal fragment of the P. falciparum CSP by sera from children living in a malaria-endemic region was associated with protection from disease. These results underline the potential of using such peptides as malaria vaccine candidates. RESUME L'intérêt de développer des stratégies d'intervention contre la malaria ciblant le stade pré-erythrocytaire a été alimenté par des études qui montrent qu'il est possible d'obtenir une immunité par l'injection de sporozoites irradiés. Les médicaments et les insecticides anti-paludiques ont été largement utilisés pour contrôler la maladie, mais dans l'espoir de développer une stratégie d'intervention plus rentable, le développement de vaccins a été placé au centre des recherches actuelles contre la malaria. A l'heure actuelle, il n'existe aucun vaccin contre la malaria. L'immunité induite par les sporozoites irradiés est due à l'effet combiné d'anticorps et de cellules T qui agissent contre les antigènes du stade hépatique dont le plus abondant est la protéine circumsporozoite (CSP). Beaucoup de formulations de vaccin visent à imiter l'immunité induite par les sporozoites irradiés. Cependant, les mécanismes par lesquels les anticorps et les cellules T sont génerés après infection par les sporozoites ou après immunisation avec des formulations de vaccin ne sont pas bien compris. La première partie de ce travail a visé à déterminer la capacité de hépatocytes primaires provenant de souris BALB/c à "processer" et à présenter des peptides dérivés de la CSP, après infection par des sporozoites de Plasmodium berghei. Nous avons montré que in vitro, les hépatocytes infectés et ceux traversés par les sporozoites pendant leur migration étaient capables de "processer" et de présenter la CSP aux cellules T CD8+ spécifiques. La voie de présentation implique le protéasome, les protéases de type aspartique et le transport à travers un compartiment post-reticulum endoplasmique. Ces résultats suggèrent que in vivo, les hépatocytes infectés contribuent à l'induction et à l'expansion d'une réponse immunitaire spécifique aux cellules T. Dans la deuxième partie, nous avons caractérisé les réponses anticorps chez les souris de la souche CB6F1 face aux peptides N- et C-terminaux des protéines circumsporozoites de Plasmodium berghei et Plasmodium falciparum. Les souris ont été immunisées avec les peptides individuellement ou en combinaison. Les peptides utilisés étaient immunogéniques chez les souris, et les anticorps produits pouvaient reconnaître la protéine CSP native à la surface des sporozoites. In vitro, les sera contre les peptides N-teminaux et les combinaisons étaient capables d'inhiber l'invasion de hépatocytes par les sporozoites. In vivo, plus de souris immunisées avec les peptides individuels de la CSP de P. berghei étaient protégées contre la malaria que les souris immunisées avec une combinaison de peptides N- et C-terminaux. De plus, les peptides N-terminaux de la CSP de P. falciparum ont été reconnus par les sera de personnes vivant dans des régions endémiques pour la malaria. Il est intéressant de voir que la reconnaissance d'un peptide N-terminal de P. falciparum par des sera d'enfants habitant dans des régions endémiques était associé à la protection contre la maladie. Ces résultats soulignent le potentiel de ces peptides comme candidats-vaccin contre la malaria.

Face Recognition using Probabilistic Neural Networks

In this paper we address the problem of face detection and recognition of grey scale frontal view images. We propose a face recognition system based on probabilistic neural networks (PNN) architecture. The system is implemented using voronoi/ delaunay tessellations and template matching. Images are segmented successfully into homogeneous regions by virtue of voronoi diagram properties. Face verification is achieved using matching scores computed by correlating edge gradients of reference images. The advantage of classification using PNN models is its short training time. The correlation based template matching guarantees good classification results

Face Recognition using Probabilistic Neural Networks

n this paper we address the problem of face detection and recognition of grey scale frontal view images. We propose a face recognition system based on probabilistic neural networks (PNN) architecture. The system is implemented using voronoi/ delaunay tessellations and template matching. Images are segmented successfully into homogeneous regions by virtue of voronoi diagram properties. Face verification is achieved using matching scores computed by correlating edge gradients of reference images. The advantage of classification using PNN models is its short training time. The correlation based template matching guarantees good classification results.

Face processing in humans is compatible with a simple shape-based model of vision

Understanding how the human visual system recognizes objects is one of the key challenges in neuroscience. Inspired by a large body of physiological evidence (Felleman and Van Essen, 1991; Hubel and Wiesel, 1962; Livingstone and Hubel, 1988; Tso et al., 2001; Zeki, 1993), a general class of recognition models has emerged which is based on a hierarchical organization of visual processing, with succeeding stages being sensitive to image features of increasing complexity (Hummel and Biederman, 1992; Riesenhuber and Poggio, 1999; Selfridge, 1959). However, these models appear to be incompatible with some well-known psychophysical results. Prominent among these are experiments investigating recognition impairments caused by vertical inversion of images, especially those of faces. It has been reported that faces that differ "featurally" are much easier to distinguish when inverted than those that differ "configurally" (Freire et al., 2000; Le Grand et al., 2001; Mondloch et al., 2002) ??finding that is difficult to reconcile with the aforementioned models. Here we show that after controlling for subjects' expectations, there is no difference between "featurally" and "configurally" transformed faces in terms of inversion effect. This result reinforces the plausibility of simple hierarchical models of object representation and recognition in cortex.

The social relations of business representation and devolved governance in the United Kingdom

The reform of regional governance in the United Kingdom has been, in part, premised on the notion that regions provide new territories of action in which cooperative networks between business communities and state-agencies can be established. Promoting business interests is seen as one mechanism for enhancing the economic competitiveness and performance of 'laggard' regions. Yet, within this context of change, business agendas and capacities are often assumed to exist 'out there, as a resource waiting to be tapped by state institutions. There is little recognition that business organisations' involvement in networks of governance owes much to historical patterns and practices of business representation, to the types of activities that exist within the business sector, and to interpretations of their own role and position within wider policymaking and implementation networks. This paper, drawing on a study of business agendas in post-devolution Scotland, demonstrates that in practice business agendas are highly complex. Their formation in any particular place depends on the actions of reflexive agents, whose perspectives and capacities are shaped by the social, economic, and political contexts within which they are operating. As such, any understanding of business agendas needs to identify the social relations of business as a whole, rather than assuming away such complexities.