Simulation Environment for Multi-Robot Cooperative 3D Target Perception

4th International Conference, SIMPAR 2014, Bergamo, Italy, October 20-23, 2014

Uncertainty based Multi-Robot Cooperative Triangulation

The paper presents a multi-robot cooperative framework to estimate the 3D position of dynamic targets, based on bearing-only vision measurements. The uncertainty of the observation provided by each robot equipped with a bearing-only vision system is effectively addressed for cooperative triangulation purposes by weighing the contribution of each monocular bearing ray in a probabilistic manner. The envisioned framework is evaluated in an outdoor scenario with a team of heterogeneous robots composed of an Unmanned Ground and Aerial Vehicle.

Towards Human-like Bimanual Movements in Anthropomorphic Robots: A Nonlinear Optimization Approach

Previously we have presented a model for generating human-like arm and hand movements on an unimanual anthropomorphic robot involved in human-robot collaboration tasks. The present paper aims to extend our model in order to address the generation of human-like bimanual movement sequences which are challenged by scenarios cluttered with obstacles. Movement planning involves large scale nonlinear constrained optimization problems which are solved using the IPOPT solver. Simulation studies show that the model generates feasible and realistic hand trajectories for action sequences involving the two hands. The computational costs involved in the planning allow for real-time human robot-interaction. A qualitative analysis reveals that the movements of the robot exhibit basic characteristics of human movements.

Robots and humans as co-workers? The human-centred perspective of work with autonomous systems

The design of work organisation systems with automated equipment is facing new challenges and the emergence of new concepts. The social aspects that are related with new concepts on the complex work environments (CWE) are becoming more relevant for that design. The work with autonomous systems implies options in the design of workplaces. Especially that happens in such complex environments. The concepts of “agents”, “co-working” or “human-centred technical systems” reveal new dimensions related to human-computer interaction (HCI). With an increase in the number and complexity of those human-technology interfaces, the capacities of human intervention can become limited, originating further problems. The case of robotics is used to exemplify the issues related with automation in working environments and the emergence of new HCI approaches that would include social implications. We conclude that studies on technology assessment of industrial robotics and autonomous agents on manufacturing environment should also focus on the human involvement strategies in organisations. A needed participatory strategy implies a new approach to workplaces design. This means that the research focus must be on the relation between technology and social dimensions not as separate entities, but integrated in the design of an interaction system.

Interaction of an opportunistic fungus Purpureocillium lilacinum with human macrophages and dendritic cells

IntroductionPurpureocillium lilacinum is emerging as a causal agent of hyalohyphomycosis that is refractory to antifungal drugs; however, the pathogenic mechanisms underlying P. lilacinum infection are not understood. In this study, we investigated the interaction of P. lilacinum conidia with human macrophages and dendritic cells in vitro.MethodsSpores of a P. lilacinum clinical isolate were obtained by chill-heat shock. Mononuclear cells were isolated from eight healthy individuals. Monocytes were separated by cold aggregation and differentiated into macrophages by incubation for 7 to 10 days at 37°C or into dendritic cells by the addition of the cytokines human granulocyte-macrophage colony stimulating factor and interleukin-4. Conidial suspension was added to the human cells at 1:1, 2:1, and 5:1 (conidia:cells) ratios for 1h, 6h, and 24h, and the infection was evaluated by Giemsa staining and light microscopy.ResultsAfter 1h interaction, P. lilacinum conidia were internalized by human cells and after 6h contact, some conidia became inflated. After 24h interaction, the conidia produced germ tubes and hyphae, leading to the disruption of macrophage and dendritic cell membranes. The infection rate analyzed after 6h incubation of P. lilacinumconidia with cells at 2:1 and 1:1 ratios was 76.5% and 25.5%, respectively, for macrophages and 54.3% and 19.5%, respectively, for cultured dendritic cells.ConclusionsP. lilacinum conidia are capable of infecting and destroying both macrophages and dendritic cells, clearly demonstrating the ability of this pathogenic fungus to invade human phagocytic cells.

Hand gesture recognition for human computer interaction : a comparative study of different image features

Hand gesture recognition for human computer interaction, being a natural way of human computer interaction, is an area of active research in computer vision and machine learning. This is an area with many different possible applications, giving users a simpler and more natural way to communicate with robots/systems interfaces, without the need for extra devices. So, the primary goal of gesture recognition research is to create systems, which can identify specific human gestures and use them to convey information or for device control. For that, vision-based hand gesture interfaces require fast and extremely robust hand detection, and gesture recognition in real time. In this study we try to identify hand features that, isolated, respond better in various situations in human-computer interaction. The extracted features are used to train a set of classifiers with the help of RapidMiner in order to find the best learner. A dataset with our own gesture vocabulary consisted of 10 gestures, recorded from 20 users was created for later processing. Experimental results show that the radial signature and the centroid distance are the features that when used separately obtain better results, with an accuracy of 91% and 90,1% respectively obtained with a Neural Network classifier. These to methods have also the advantage of being simple in terms of computational complexity, which make them good candidates for real-time hand gesture recognition.

Generic system for human-computer gesture interaction: applications on sign language recognition and robotic soccer refereeing

Hand gestures are a powerful way for human communication, with lots of potential applications in the area of human computer interaction. Vision-based hand gesture recognition techniques have many proven advantages compared with traditional devices, giving users a simpler and more natural way to communicate with electronic devices. This work proposes a generic system architecture based in computer vision and machine learning, able to be used with any interface for human-computer interaction. The proposed solution is mainly composed of three modules: a pre-processing and hand segmentation module, a static gesture interface module and a dynamic gesture interface module. The experiments showed that the core of visionbased interaction systems could be the same for all applications and thus facilitate the implementation. For hand posture recognition, a SVM (Support Vector Machine) model was trained and used, able to achieve a final accuracy of 99.4%. For dynamic gestures, an HMM (Hidden Markov Model) model was trained for each gesture that the system could recognize with a final average accuracy of 93.7%. The proposed solution as the advantage of being generic enough with the trained models able to work in real-time, allowing its application in a wide range of human-machine applications. To validate the proposed framework two applications were implemented. The first one is a real-time system able to interpret the Portuguese Sign Language. The second one is an online system able to help a robotic soccer game referee judge a game in real time.

Generic system for human-computer gesture interaction

Hand gestures are a powerful way for human communication, with lots of potential applications in the area of human computer interaction. Vision-based hand gesture recognition techniques have many proven advantages compared with traditional devices, giving users a simpler and more natural way to communicate with electronic devices. This work proposes a generic system architecture based in computer vision and machine learning, able to be used with any interface for humancomputer interaction. The proposed solution is mainly composed of three modules: a pre-processing and hand segmentation module, a static gesture interface module and a dynamic gesture interface module. The experiments showed that the core of vision-based interaction systems can be the same for all applications and thus facilitate the implementation. In order to test the proposed solutions, three prototypes were implemented. For hand posture recognition, a SVM model was trained and used, able to achieve a final accuracy of 99.4%. For dynamic gestures, an HMM model was trained for each gesture that the system could recognize with a final average accuracy of 93.7%. The proposed solution as the advantage of being generic enough with the trained models able to work in real-time, allowing its application in a wide range of human-machine applications.

Global vs. local nonlinear optimization techniques for human-like movement of an anthropomorphic robot

In this paper a comparison between using global and local optimization techniques for solving the problem of generating human-like arm and hand movements for an anthropomorphic dual arm robot is made. Although the objective function involved in each optimization problem is convex, there is no evidence that the admissible regions of these problems are convex sets. For the sequence of movements for which the numerical tests were done there were no significant differences between the optimal solutions obtained using the global and the local techniques. This suggests that the optimal solution obtained using the local solver is indeed a global solution.

Towards human-like bimanual movements in anthropomorphic robots: a nonlinear optimization approach

Previously we have presented a model for generating human-like arm and hand movements on an unimanual anthropomorphic robot involved in human-robot collaboration tasks. The present paper aims to extend our model in order to address the generation of human-like bimanual movement sequences which are challenged by scenarios cluttered with obstacles. Movement planning involves large scale nonlinear constrained optimization problems which are solved using the IPOPT solver. Simulation studies show that the model generates feasible and realistic hand trajectories for action sequences involving the two hands. The computational costs involved in the planning allow for real-time human robot-interaction. A qualitative analysis reveals that the movements of the robot exhibit basic characteristics of human movements.

Hand gesture recognition system based in computer vision and machine learning: Applications on human-machine interaction

Tese de Doutoramento em Engenharia de Eletrónica e de Computadores

Multimodal automatic user disposition recognition in human-machine interaction

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

Photodynamic drug delivery enhancement in tumours does not depend on leukocyte-endothelial interaction in a human mesothelioma xenograft model.

OBJECTIVES: The pre-treatment of tumour neovessels by low-level photodynamic therapy (PDT) improves the distribution of concomitantly administered systemic chemotherapy. The mechanism by which PDT permeabilizes the tumour vessel wall is only partially known. We have recently shown that leukocyte-endothelial cell interaction is essential for photodynamic drug delivery to normal tissue. The present study investigates whether PDT enhances drug delivery in malignant mesothelioma and whether it involves comparable mechanisms of actions. METHODS: Human mesothelioma xenografts (H-meso-1) were grown in the dorsal skinfold chambers of 28 nude mice. By intravital microscopy, the rolling and recruitment of leukocytes were assessed in tumour vessels following PDT (Visudyne(®) 400 μg/kg, fluence rate 200 mW/cm(2) and fluence 60 J/cm(2)) using intravital microscopy. Likewise, the distribution of fluorescently labelled macromolecular dextran (FITC-dextran, MW 2000 kDa) was determined after PDT. Study groups included no PDT, PDT, PDT plus a functionally blocking anti-pan-selectin antibody cocktail and PDT plus isotype control antibody. RESULTS: PDT significantly enhanced the extravascular accumulation of FITC-dextran in mesothelioma xenografts, but not in normal tissue. PDT significantly increased leukocyte-endothelial cell interaction in tumour. While PDT-induced leukocyte recruitment was significantly blunted by the anti-pan-selectin antibodies in the tumour xenograft, this manipulation did not affect the PDT-induced extravasation of FITC-dextran. CONCLUSIONS: Low-level PDT pre-treatment selectively enhances the uptake of systemically circulating macromolecular drugs in malignant mesothelioma, but not in normal tissue. Leukocyte-endothelial cell interaction is not required for PDT-induced drug delivery to malignant mesothelioma.

Anionic Sites, Fucose Residues and Class I Human Leukocyte Antigen Fate During Interaction of Toxoplasma gondii with Endothelial Cells

Toxoplasma gondii invades and proliferates in human umbilical vein endothelial cells where it resides in a parasitophorous vacuole. In order to analyze which components of the endothelial cell plasma membrane are internalized and become part of the parasitophorous vacuole membrane, the culture of endothelial cells was labeled with cationized ferritin or UEA I lectin or anti Class I human leukocytte antigen (HLA) before or after infection with T. gondii. The results showed no cationized ferritin and UEA I lectin in any parasitophorous vacuole membrane, however, the Class I HLA molecule labeling was observed in some endocytic vacuoles containing parasite until 1 h of interaction with T. gondii. After 24 h parasite-host cell interaction, the labeling was absent on the vacuolar membrane, but presents only in small vesicles near parasitophorous vacuole. These results suggest the anionic site and fucose residues are excluded at the time of parasitophorous vacuole formation while Class I HLA molecules are present only on a minority of Toxoplasma-containig vacuoles.

The interaction of human pathogenic arena viruses with the host cell

SummaryResearch projects presented in this thesis aimed to investigate two major aspects of the arenaviruses life cycle in the host cell: viral entry and the biosynthesis of the viral envelope glycoprotein.Old World arenaviruses (OWAV), such as Lassa virus (LASV) and lymphocytic choriomeningitis virus (LCMV), attach to the cell by binding to their receptor, alpha-dystroglycan. Virions are then internalized by a largely unknown pathway of endocytosis and delivered to the late endosome/lysosome where fusion occurs at low pH. In the major project of my thesis, we sought to identify cellular factors involved in OWAV cell entry. Our work indicates that OWAV cell entry requires microtubular transport and a functional multivesicular body (MVB) compartment. Infection indeed depends on phosphatidyl inositol 3-kinase (PI3K) activity and lysobisphosphatidic acid (LBPA), a lipid found in membranes of intraluminal vesicles (ILVs) of the MVB. We further found a requirement of factors that are part of the endosomal sorting complex required for transport (ESCRT), involved in the formation of ILVs. This suggests an ESCRT-mediated sorting of virus- receptor complex during the entry process.During viral replication, biosynthesis of viral glycoprotein takes place in the endoplasmic reticulum (ER) of the host cell. When protein load exceeds the folding capacity of the ER, the accumulation of unfolded proteins is sensed by three ER resident proteins, activating transcription factor 6 (ATF6), inositol-requiring enzyme 1 (IRE1) and PKR-like ER kinase (PERK), whose signaling induces the cellular unfolded protein response (UPR). Our results indicate that acute LCMV infection transiently induces the activation of the ATF6 branch of the UPR, whereas the PERK, and IRE1 axis of UPR are neither triggered nor blocked during infection. Our data also demonstrate that activation of ATF6 pathway is required for optimal viral replication during acute infection.The formation of the mature, fusion-active form of arenaviruses glycoproteins requires proteolytic cleavage mediated by the cellular protease subtilisin kexin isozyme-1 (SKI-l)/site-l protease (SIP). We show that targeting the SKI-1/S1P enzymatic activity with specific inhibitors is a powerful strategy to block arenaviruses productive infection. Moreover, characterization of protease function highlights differences in processing between cellular and viral substrates, opening new possibilities in term of drug development against human pathogenic arenaviruses.RésuméLes projets de recherche présentés dans cette thèse visaient à étudier deux aspects du cycle de vie des arenavirus: l'entrée du virus dans la cellule hôte et la biosynthèse de la glycoprotéine durant la réplication virale.Les arenavirus du vieux monde (OWAV), tels que le virus de Lassa (LASV) et le virus de la chorioméningite lymphocytaire (LCMV) s'attachent à la cellule hôte en se liant à leur récepteur, l'alpha-dystroglycane. Les virions sont ensuite intemalisés par une voie d'endocytose inconnue et livrés à l'endosome tardif/lysosome, où le pH acide permet la fusion entre l'enveloppe virale et la membrane du compartiment. Le projet principal de ma thèse consistait à identifier les facteurs cellulaires impliqués dans l'entrée des OWAV dans la cellule hôte. Nos résultats indiquent que l'entrée des OWAV nécessite le transport microtubulaire et la présence d'un corps multivésiculaire (MVB) fonctionnel. L'infection dépend en effet de l'activité de phosphatidyl inositol 3-kinase (PI3K) et de lysobisphosphatidic acid (LBPA), un lipide présent dans les membranes des vésicules intraluminales (ILVs) du MVB. Nous avons également trouvé l'implication de facteurs constituant l'endosomal sorting complex required for sorting (ESCRT) qui joue un rôle dans la formation des ILVs. Ces donnés suggèrent l'incorporation du complexe virus-récepteur dans des ILVs durant le processus d'entrée.Lors de la réplication virale, la biosynthèse de la glycoprotéine virale a lieu dans le réticulum endoplasmique (ER) de la cellule hôte. Lorsque la charge de protéines nouvellement synthétisées excède la capacité de pliage des protéines dans le ER, l'accumulation de protéines mal pliées est détectée par trois facteurs: activating transcription factor 6 (ATF6), inositol-requiring enzyme 1 (IRE1) et PKR-like ER kinase (PERK). Leur signalisation constitue la réponse cellulaire face aux protéines mal pliées (UPR). Nos résultats montrent que l'infection aiguë avec LCMV induit transitoirement l'activation de la voie de signalisation ATF6 alors que les axes PERK et IRE1 de l'UPR ne sont ni induits ni bloqués pendant l'infection. Nos données prouvent également que l'activation de la voie ATF6 est nécessaire à une réplication virale optimale lors de l'infection aiguë avec LCMV.La maturation des glycoprotéines des arenavirus nécessite un clivage protéolytique par la protéase cellulaire subtilisin kexin isozyme-1 (SKI-l)/site-l protease (SIP). Nous avons démontré que le ciblage de l'activité enzymatique de SKI-1/SIΡ avec des inhibiteurs spécifiques est une stratégie prometteuse pour bloquer l'infection par les arenavirus. La caractérisation du mécanisme d'action de la protéase a, par ailleurs, révélé des différences au niveau du clivage entre les substrats cellulaires et viraux, ce qui ouvre de nouvelles perspectives en terme de développement de médicaments contre les arenavirus pathogènes pour l'homme.