No description available.

The aim of the experiment was to evaluate the effectiveness of selected specialty fertilizers on the quality of nursery plants. The experiment was established in May 2013 and lasted for one growing season. It compares five options. To enrich the growing media nutrients were applied NPK fertilizer with micronutrients, Multicote 8 M, Osmocote 5-6 M and T Nutricote 100 for fertilizing during the growing season Kristalon Vega, Kristalon Gena, Yara Mila Complex and Vitality Complex. The experiment included 1 000 plants botanical species Cotoneaster dammeri SCHNEID. Endpoint was the length of the longest growth, the number of shoots, root system volume and the number of dead plants. Best results were achieved variant K, which was used for a combination of NPK fertilizer with micronutrients, Kristalon and Yara Mila Complex. Followed by variant fertilized combination products Multicote 8 M, Kristalon and Vitality Complex. Worse results were achieved by a combination of variants of fertilization Kristalon. The substrate for these variants were enriched fertilizers Multicom 8 M, Osmocote 5-6 M and Nutricote T 100, where the worst-hit variant using Nutricote T 100. Highest incidence of dead plants was in the variant K.

We consider the problem of enabling a robot to efficiently obtain a dense haptic map of its visible surroundings Using the complementary properties of vision and tactile sensing. Our approach assumes that visible surfaces that look similar to one another are likely to have similar haptic properties. In our previous work, we introduced an iterative algorithm that enabled a robot to infer dense haptic labels across visible surfaces in an RGB-D image when given a sequence of sparse haptic labels. In this work, we describe how dense conditional random fields (CRFs) can be applied to this same problem and present results from evaluating a dense CRF’s performance in simulated trials with idealized haptic labels. We evaluated our method using several publicly available RGB-D image datasets with indoor cluttered scenes pertinent to robot manipulation. In these simulated trials, the dense CRF substantially outperformed our previous algorithm by correctly assigning haptic labels to an average of 93% (versus 76% in our previous work) of all object pixels in an image given the highest number of contact points per object. Likewise, the dense CRF correctly assigned haptic labels to an average of 81% (versus 63% in our previous work) of all object pixels in an image given a low number of contact points per object. We compared the performance of dense CRF using uniform prior with a dense CRF using prior obtained from the visible scene using a Fully Convolutional Network trained for visual material recognition. The use of the convolutional network further improves the performance of the algorithm. We also performed experiments with the humanoid robot DARCI reaching in a cluttered foliage environment while using our algorithm to create a haptic map. The algorithm correctly assigned the label to 82.52% of the scenes with trunks and leaves after 10 reaches into the environment.

This work consists of an investigation into the properties of unknown words in HPSG, and in particular into the phenomenon of multi-word unknown expressions consisting of multiple unknown words in a sequence. The work presented consists first of a study determining the relative frequency of multi-word unknown expressions, and then a survey of the efficacy of a variety of techniques for handling these expressions. The techniques presented consist of modified versions of techniques from the existing unknown-word prediction literature as well as novel techniques, and they are evaluated with a specific concern for how they fare in the context of sentences with many unknown words and long unknown sequences. / text

Stress is defined as a challenge to homeostatic equilibrium by physical or psychological events, generating a coping response consisting of central and peripheral changes, with the aim of exerting control over the threatening events. Corticotropin-releasing factor (CRF) is well known as a hypothalamic factor which controls the hypothalamic-pituitary-adrenocortical (HPA) axis during basal activity and stress. CRF also serves a neurotransmitter function in the brain, where it is implicated in a range of stress-related behaviours. The measurement of local cerebral glucose utilisation (LCGU) using radiolabelled 2-deoxy-D-glucose (2DG) provides an estimate of cellular activity in the brain. 2DG competes with glucose in its metabolic pathway, but is not fully metabolised, instead accumulating within astrocytes where it can be quantified. After consideration of available modifications to the LCGU technique, the effect of manipulating the CRF system on LCGU was studied, in order to test the hypothesis that CRF and other endogenously expressed CRF-related peptides would induce different patterns of LCGU, and to examine the involvement of CRF receptors in any response. The CRF1 receptor has been implicated in the mediation of stress- and anxiety-related behaviour, while recent evidence has suggested a role for CRF2 in mediating the delayed effects of stress, although it has previously been postulated that CRF2 may be involved in the attenuation of stress-related behaviour. CRF and the endogenous CRF-related peptide Urocortin 1 both induced increases in LCGU in a number of brain regions associated with the CRF system, with concomitant activation of the HPA axis. CRF induced increases in LCGU in the dissected hypothalamus, thalamus, cerebellum and hippocampus, while Urocortin 1 induced a significant increase in LCGU in a dissected hindbrain region, with trend-like effects in frontal cortex and hippocampus. These regions contain components of the CRF system, or receive projections from regions involved in the CRF system, and have been implicated in stress-related function. The effects of CRF on LCGU appear to be mediated by the CRF2 receptor, as they were abolished by the selective CRF2 antagonist antisauvagine-30, but persisted in mice lacking CRF1 and were unaffected by a selective CRF1 antagonist. However, neither of the endogenous CRF-related peptides selective for CRF2, Urocortin 2 and Stresscopin, affected LCGU, which may indicate ligand-specific effects within the CRF system. In contrast to the effects of CRF, restraint stress reduced LCGU, while activating the HPA axis, and this response was unaffected by a selective CRF1 antagonist. This data suggests that the role of CRF receptors in restraint-induced LCGU changes may be overshadowed by effects on other neurotransmitter systems. These studies support the hypothesis that CRF and other endogenously expressed CRF-related peptides would induce different patterns of LCGU, and highlight the involvement of particular brain regions in the response to CRF receptor stimulation. Furthermore, these studies provide evidence suggesting ligand-specific effects within the CRF system.

In order to interact with the real world, humans need to perform several tasks such as object detection, pose estimation, motion estimation and distance estimation. These tasks are all part of scene understanding and are fundamental tasks of computer vision. Depth estimation received unprecedented attention from the research community in recent years due to the growing interest in its practical applications (ie robotics, autonomous driving, etc.) and the performance improvements achieved with deep learning. In fact, the applications expanded from the more traditional tasks such as robotics to new fields such as autonomous driving, augmented reality devices and smartphones applications. This is due to several factors. First, with the increased availability of training data, bigger and bigger datasets were collected. Second, deep learning frameworks running on graphical cards exponentially increased the data processing capabilities allowing for higher precision deep convolutional networks, ConvNets, to be trained. Third, researchers applied unsupervised optimization objectives to ConvNets overcoming the hurdle of collecting expensive ground truth and fully exploiting the abundance of images available in datasets. This thesis addresses several proposals and their benefits for unsupervised depth estimation, i.e., (i) learning from resynthesized data, (ii) adversarial learning, (iii) coupling generator and discriminator losses for collaborative training, and (iv) self-improvement ability of the learned model. For the first two points, we developed a binocular stereo unsupervised depth estimation model that uses reconstructed data as an additional self-constraint during training. In addition to that, adversarial learning improves the quality of the reconstructions, further increasing the performance of the model. The third point is inspired by scene understanding as a structured task. A generator and a discriminator joining their efforts in a structured way improve the quality of the estimations. Our intuition may sound counterintuitive when cast in the general framework of adversarial learning. However, in our experiments we demonstrate the effectiveness of the proposed approach. Finally, self-improvement is inspired by estimation refinement, a widespread practice in dense reconstruction tasks like depth estimation. We devise a monocular unsupervised depth estimation approach, which measures the reconstruction errors in an unsupervised way, to produce a refinement of the depth predictions. Furthermore, we apply knowledge distillation to improve the student ConvNet with the knowledge of the teacher ConvNet that has access to the errors.

The present set of experiments were designed to examine the role of histamine, and its interaction with corticotropin releasing factor (CRF) and bombesin (BM) in food intake regulation of chickens. The hypothesis being tested was as follows: One component of the neuroregulation of food intake involves histaminergic activity in the hypothalamus, acting on either H1 or H2 receptors, how these receptors interact with CRF neurons and if BM elicits its effects on feeding through CRF release. Single Comb White Leghorn (SCWL) and broiler cockerels were utilized for these experiments. Birds were stereotaxically implanted with a 23-gauge thin-walled stainless steel guide cannula, and were provided a mash diet and water for ad libitum consumption. All compounds were infused into the right lateral ventricle. Effects were monitored at 15-minute intervals through three hours postinjection. Experiment 1 examined the effects of intracereboventricular (ICV) injections of histamine (HA) and two HA antagonists, the H1 receptor antagonist chloropheneramine maleate (CM) and H2 receptor antagonist cimetidine (CIM), on food and water consumption and body temperature. Histamine was infused using 0, 25, 50, and 100 µg per 10 µl of artificial cerebrospinal fluid (aCSF). Histamine significantly decreased food and water consumption (P< 0.05) over the three hour observation period in a dose-dependent manner. Histamine was then infused to observe if the decrease in water intake was dependent upon the decrease in food intake. Birds were not allowed access to feed during this experiment. Water intake was not affected by HA in either SCWL or broilers when food was not available. To observe the effects of HA on thermoregulation, HA was infused using the same dosages and body temperature recorded for three hours. Histamine produced hypothermia at a dose of 25 µg in SCWL cockerels, with a quadratic trend at 165 and 180 min. Broiler cockerels did not show hypothermia, but rather a constant hyperthermia compared to the control with a quadratic trend throughout the latter part of the experiment. The last phase of the first set of experiments, birds were pretreated with either CM or CIM (100 µg/10 µl aCSF) followed by HA. When the birds were pretreated with either CM or CIM, the hypophagic responses to HA were attenuated. The pair of experiments that utilized H1 and H2 receptors demonstrated that these receptors are involved in the neural regulation of food intake. These experiments also demonstrated that the aphagic effects of HA on food intake can be blocked with the pretreatment of antihistaminics affecting both H1 and H2 receptors. In Experiment 2, studies were conducted to determine if neuronal CRF elicited its effects on feeding through the release of HA. Birds were infused with 0 or 20 µg CRF and either 0 or 100 µg of CM or CIM. CRF decreased food and water intake in both SCWL and broiler cockerels. When birds were pretreated with CM, the hypophagic responses to CRF were attenuated. When birds were pretreated with CIM, the hypophagic responses of CRF were attenuated in broiler cockerels; this response was not seen in SCWL cockerels. Water intake followed a similar pattern. It was concluded that, contrary to studies showing that HA causes the release of CRF in other species, CRF may cause the release of HA in chickens. Experiment 3 was designed to investigate whether bombesin (BM) elicited its effects on feeding through the release of CRF. Birds were infused with either, 0 or 0.5 µg BM, 0 or 5 µg aCRF (9-41) (CRF antagonist), or a combination of both. These compounds were infused to test whether the effects of BM could be blocked with the pretreatment of anticorticotropics. Food and water consumption were significantly decreased (P< 0.05) with the infusion of BM in both SCWL and broiler cockerels. Food intake was not affected with the infusion of aCRF in SCWL or broilers cockerels. However, water consumption was increased when birds were given ICV injections of aCRF. When birds were pretreated with aCRF, the anorexigenic and adipsic effects of BM were attenuated. It was concluded that BM elicits its effects on feeding through the release of CRF. These results also demonstrate that the aphagic effects of BM could be blocked with the pretreatment of anticorticotropics. / Master of Science

Dans le contexte de la vision par ordinateur cette thèse étudie le problème d’appariement d’images dans le cadre de la télédétection pour la géologie. Plus précisément, nous disposons dans ce travail de deux images de la même scène géographique, mais acquises à partir de deux points de vue différents et éventuellement à un autre moment. La tâche d’appariement est d'associer à chaque pixel de la première image un pixel de la seconde image.Bien que ce problème soit relativement facile pour les êtres humains, il reste difficile à résoudre par un ordinateur. De nombreuses approches pour traiter cette tâche ont été proposées. Les techniques les plus prometteuses formulent la tâche comme un problème d'optimisation numérique. Malheureusement, le nombre d'inconnues ainsi que la nature de la fonction à optimiser rendent ce problème extrêmement difficile à résoudre. Cette thèse étudie deux approches avec un schéma multi-échelle pour résoudre le problème numérique sous-jacent / This thesis studies the computer vision problem of image registration in the context of geological remote sensing surveys. More precisely we dispose in this work of two images picturing the same geographical scene but acquired from two different view points and possibly at a different time. The task of registration is to associate to each pixel of the first image its counterpart in the second image.While this problem is relatively easy for human-beings, it remains an open problem to solve it with a computer. Numerous approaches to address this task have been proposed. The most promising techniques formulate the task as a numerical optimization problem. Unfortunately, the number of unknowns along with the nature of the objective function make the optimization problem extremely difficult to solve. This thesis investigates two approaches along with a coarsening scheme to solve the underlying numerical problem

Les systèmes de télécommunications actuels nécessitent des filtres passe bande fonctionnant à des fréquences comprises entre 1 GHz et 10 GHz pour les systèmes les plus répandus, notamment la téléphonie mobile. Les filtres actuels sont reportés, donc non intégrés sur silicium. Ils présentent certains inconvénients : coût, place occupée, incompatibilité avec les technologies silicium…Une solution consiste à utiliser des résonateurs à ondes de volume, plus communément appelés BAW (Bulk Acoustic Waves). Ils présentent l'avantage d'être intégrables sur silicium. De nouvelles architectures nommées CRF (Coupled Resonator Filter) font appel à des résonateurs à ondes de volume (BAW) mais aussi à des structures utilisant des couplages acoustiques entre différentes couches. L'objectif du travail proposé est de modéliser les structures actuelles et en cours d'études pour obtenir des modèles de type circuits électriques. Ces modèles seront validés par des mesures effectuées au laboratoire sur des résonateurs BAW et sur des filtres qui sont réalisés par nos partenaires. La première partie des études portait sur le comportement RF sous faible et forte puissances, suivi du développement d’un modèle large bande simulé sous ADS décrivant le comportement des filtres BAW sous faible et forte puissances. Une optimisation du temps de mesure en fréquence est effectuée pour réduire le temps du test RF. La seconde partie de la thèse est orientée vers le développement d'une nouvelle méthode de test pseudo-temporelle des filtres hyperfréquences qui consiste à mesurer directement leur impact sur un signal numérique grâce au paramètre «EVM» (Error Vector Magnitude). Ce paramètre est relié au BER et nos travaux montrent qu’il permet également de retrouver en partie les paramètres S et détecter les filtres défaillants à partir d’une seule mesure. Cette nouvelle technique permettant le test de filtres à partir d’un seul point de mesure permet de réduire le temps et le coût de caractérisation à des fins industrielles. Ce travail s'est déroulé dans un cadre de collaborations avec le LETI et STMicroelectronics au sein du projet FAST labellisé par le pôle MINALOGIC. / The telecommunication systems require pass band filters operating between 1 GHz and 10 GHz for most of the popular radio communication system including the mobile phones. The current filters that are not integrated on silicon present some inconveniences such as cost, surface occupied and incompatibility with silicon technology. The solution consists on using Bulk Acoustic Wave resonator called BAW that has the advantage of being integrated on silicon. Innovative architectures called CRF (Coupled Resonator Filter) are based on BAW resonator and use acoustic coupling between different layers. To understand the design and the functioning of these structures, finite element software can be used which requires significant computing time. The objective of our work is modelling the existing and coming structures with models made of electrical circuits. These models will be validated on BAW filters designed by our partners (CEA-LETI) and measured in our laboratory. In the first part, we develop an electro-acoustical wide band model of BAW and CRF filters simulated with ADS software for low and high RF powers. Optimization of measurement in the frequency domain is made to reduce the time of RF test. The second part of the thesis is directed towards the development of a new method of pseudo-time domain test of RF filters, which consists on measuring directly their impact on a digital signal with the parameter «EVM» (Error Vector Magnitude) that is related to BER. Our work shows that we can also find some of S parameters and detect the defected filters from a single measurement extraction. This new technique of filter testing with a single measurement point allows reducing time and cost of test for industrial purposes. This work takes place within collaboration with CEA-LETI and STMicroelectronics in the project FAST labelled by MINALOGIC pole

The abuse of and addiction to drugs of abuse, such as tobacco, alcohol, opioids, and illicit drugs, are growing global problems that affect the welfare of individuals and societies worldwide. The National Institute of Drug Abuse estimates the annual cost of substance abuse to be over $740 billion in costs related to drug intoxication, withdrawal and relapse. A primary challenge in the treatment of substance abuse is the tendency of users to relapse following acute or extended periods of abstinence; on average over 60% of substance abusers will return to drug use within a year of receiving treatment, many relapsing following stressful life events. Central to the successful treatment of drug addiction is understanding the cellular mechanisms by which relapse episodes occur. Current data suggest that the activation of pituitary adenylate cyclase activating peptide (PACAP) systems in the bed nucleus of the stria terminalis (BNST) is an important event underlying stress-induced reinstatement to drug-seeking in a rodent model of stress-induced relapse. In conjunction with immunohistochemistry and pharmacological treatments, we used this behavioral model of stress-induced relapse to evaluate PACAP and PACAP type-1 receptor (PAC1-R) signaling in stress-induced reinstatement to cocaine seeking. Activation of the PAC1 receptor appears to be critical to stress-induced reinstatement, as the selective PAC1-R agonist, maxadilan, produced reinstatement behaviors in the absence of stress. Moreover, BNST pretreatment with either mitogen activated protein kinase-ERK (MEK) or endocytosis inhibitors to block extracellular signal-related kinase (ERK) signaling attenuated stress-induced reinstatement. Furthermore, BNST phosphorylated ERK (pERK) expression, mediated by PAC1-R activation, is substantially potentiated in cocaine-experienced animals after stressor exposure, in a manner that is dependent on endosomal signaling and MEK activity. These data suggest that the activation of a PAC1 signaling cascade is a key event underlying stress-induced reinstatement. Furthermore, this data may suggest a permanent change in the BNST PACAP system (sensitization) following cocaine exposure.