Search

Positively imagined activities may capture visual attention due to an increase in positive value. Increasing attention toward activities, in turn, may prove useful for clinical interventions aiming to motivate behavioral engagement. Employing a within-subject experimental design, we examined the effect of positive imagery on attention using a visual probe task with concurrent eye tracking. Adults from the general population (N = 54) imagined performing activities involving visually presented objects in a positive (focusing on the positive emotional impact) or neutral (focusing on a neutral circumstance) manner. They then completed a visual probe task using picture stimuli depicting one object per type of imagery. Positive compared to neutral imagery increased self-reported behavioral motivation and biased the direction, but not the duration, of gaze toward objects associated with the imagined activities. An exploratory analysis showed a positive association between the direction bias and depressive symptoms. Our findings build on existing literature on positive imagery as a motivational amplifier by highlighting early attention as an underlying cognitive mechanism.

PurposeAlthough schema therapy has been predominantly applied to treat personality disorders, interest into its application in other clinical disorders is growing. Central to schema therapy are Early Maladaptive Schemas (EMS) and Schema Modes. Since existing EMS and Schema Modes were primarily developed in the context of personality disorders, their relevance for clinical disorders is unclear. MethodsWe conducted a systematic review of the presence of EMS and Schema Modes in clinical disorders according to DSM criteria. Per disorder, we evaluated which EMS and Schema Modes were more pronounced in comparison with clinical as well as non-clinical control groups and which EMS and Schema Modes were most highly endorsed within the disorder. ResultsAlthough evidence concerning EMS was scarce for several disorders, and only few studies on Schema Modes survived inclusion criteria, we identified meaningful relationships and patterns for EMS and Schema Modes in various clinical disorders. ConclusionsThe present review highlights the relevance of EMS and Schema Modes for clinical disorders beyond personality disorders. Depending on the theme of the representation, EMS act as vulnerabilities both across diagnoses and for specific disorders. Thus, EMS and resulting Schema Modes are potential, valuable targets for the prevention and treatment of clinical disorders.

During winter, conifers at the alpine treeline suffer dramatic losses of hydraulic conductivity, which are successfully recovered during late winter. Previous studies indicated branch water uptake to support hydraulic recovery. We analyzed water absorption and redistribution in Picea abies and Larix decidua growing at the treeline by in situ exposure of branches to δ2H-labelled water. Both species suffered high winter embolism rates (> 40–60% loss of conductivity) and recovered in late winter (L. decidua showing higher rates. In natura, melting snow might be the prime source for absorbed and redistributed water, enabling embolism repair and restoration of water reservoirs prior to the vegetation period. Pronounced water uptake in the deciduous L. decidua indicated bark to participate in the process of water absorption.

Key message The lack of elevational changes in electrical resistivity in Pinus cembra trunks indicated consistent growth and hydraulics across elevations. Though, electrical resistivity tomograms exhibited pronounced temperature-driven seasonal changes. Abstract Alpine conifers growing at high elevation are exposed to low temperatures, which may limit xylogenesis and cause pronounced seasonal changes in tree hydraulics. Electrical resistivity (ER) tomography enables minimal invasive monitoring of stems in situ. We used this technique to analyze Pinus cembra trunks along a 400 m elevational gradient up to the timberline and over seasons. Furthermore, ER data of earlywood across tree rings were compared with the respective specific hydraulic conductivity (KS), measured on extracted wood cores. ER tomograms revealed pronounced changes over the year and a strong correlation between average ER (ERmean) and air and xylem temperatures. Surprisingly, no elevational changes in ERmean, earlywood ER or KS were observed. ER data corresponded to variation in earlywood KS, which decreased from the youngest (ca. 4–5 cm2s−1 MPa−1) to the oldest tree rings (0.63 ± 0.22 cm2s−1 MPa−1). The lack of changes in ER data and earlywood KS along the study transect indicated consistent growth patterns and no major changes in structural and functional hydraulic traits across elevation. The constant decrease in earlywood KS with tree ring age throughout all elevations highlights the hydraulic relevance of the outermost tree rings in P. cembra. Seasonal measurements demonstrated pronounced temperature effects on ER, and we thus recommend a detailed monitoring of trunk temperatures for ER tomography.

Modern deep neural networks (DNNs) are achieving state-of-the-art results due to their capability to learn a faithful representation of the data they are trained on. In this chapter, we address two insufficiencies of DNNs, namely, the lack of robustness to corruptions in the data, and the lack of real-time deployment capabilities, that need to be addressed to enable their safe and efficient deployment in real-time environments. We introduce hybrid corruption-robustness focused compression (HCRC), an approach that jointly optimizes a neural network for achieving network compression along with improvement in corruption robustness, such as noise and blurring artifacts that are commonly observed. For this study, we primarily consider the task of semantic segmentation for automated driving and focus on the interactions between robustness and compression of the network. HCRC improves the robustness of the network by 8.39% absolute mean performance under corruption (mPC) on the Cityscapes dataset, and by 2.93% absolute mPC on the Sim KI-A dataset, while generalizing even to augmentations not seen by the network in the training process. This is achieved with only minor degradations on undisturbed data. Our approach is evaluated over two strong compression ratios (30% and 50%) and consistently outperforms all considered baseline approaches. Additionally, we perform extensive ablation studies to further leverage and extend existing state-of-the-art methods.

The use of deep neural networks (DNNs) in safety-critical applications like mobile health and autonomous driving is challenging due to numerous model-inherent shortcomings. These shortcomings are diverse and range from a lack of generalization over insufficient interpretability to problems with malicious inputs. Cyber-physical systems employing DNNs are therefore likely to suffer from safety concerns. In recent years, a zoo of state-of-the-art techniques aiming to address these safety concerns has emerged. This work provides a structured and broad overview of them. We first identify categories of insufficiencies to then describe research activities aiming at their detection, quantification, or mitigation. Our paper addresses both machine learning experts and safety engineers: The former ones might profit from the broad range of machine learning topics covered and discussions on limitations of recent methods. The latter ones might gain insights into the specifics of modern ML methods. We moreover hope that our contribution fuels discussions on desiderata for ML systems and strategies on how to propel existing approaches accordingly., 94 pages

Although deep neural networks (DNNs) are high-performance methods for various complex tasks, e.g., environment perception in automated vehicles (AVs), they are vulnerable to adversarial perturbations. Recent works have proven the existence of universal adversarial perturbations (UAPs), which, when added to most images, destroy the output of the respective perception function. Existing attack methods often show a low success rate when attacking target models which are different from the one that the attack was optimized on. To address such weak transferability, we propose a novel learning criterion by combining a low-level feature loss, addressing the similarity of feature representations in the first layer of various model architectures, with a cross-entropy loss. Experimental results on ImageNet and Cityscapes datasets show that our method effectively generates universal adversarial perturbations achieving state-of-the-art fooling rates across different models, tasks, and datasets. Due to their effectiveness, we propose the use of such novel generated UAPs in robustness evaluation of DNN-based environment perception functions for AVs.

Xylem conductive capacity is a key determinant of plant hydraulic function and intimately linked to photosynthesis and productivity, but can be impeded by temporary or permanent conduit dysfunctions. Here we show that persistent xylem dysfunctions in unstressed plants are frequent in Alpine dwarf shrubs and occur in various but species-specific cross-sectional patterns. Combined synchrotron micro-computed tomography (micro-CT) imaging, xylem staining, and flow measurements in saturated samples of six widespread Ericaceae species evidence a high proportion (19%-50%) of hydraulically nonfunctional xylem areas in the absence of drought stress, with regular distribution of dysfunctions between or within growth rings. Dysfunctions were only partly reversible and reduced the specific hydraulic conductivity to 1.38 to 3.57 ×10-4 m2 s-1 MPa-1 . Decommission of inner growth rings was clearly related to stem age and a higher vulnerability to cavitation of older rings, while the high proportion of nonfunctional conduits in each annual ring needs further investigations. The lower the xylem fraction contributing to the transport function, the higher was the hydraulic efficiency of conducting xylem areas. Improved understanding of the functional lifespan of xylem elements and the prevalence and nature of dysfunctions is critical to correctly assess structure-function relationships and whole-plant hydraulic strategies.

Heat injuries sustained in a fire can initiate a cascade of complex mechanisms that affect the physiology of trees after fires. Uncovering the exact physiological mechanisms and relating specific injuries to whole-plant and ecosystem functioning is the focus of intense current research. Recent studies have made critical steps forward in our understanding of tree physiological processes after fires, and have suggested mechanisms by which fire injuries may interact with disturbances such as drought, insects and pathogens. We outline a conceptual framework that unifies the involved processes, their interconnections, and possible feedbacks, and contextualizes these responses with existing hypotheses for disturbance effects on plants and ecosystems. By focusing on carbon and water as currencies of plant functioning, we demonstrate fire-induced cambium/phloem necrosis and xylem damage to be main disturbance effects. The resulting carbon starvation and hydraulic dysfunction are linked with drought and insect impacts. Evaluating the precise process relationships will be crucial for fully understanding how fires can affect tree functionality, and will help improve fire risk assessment and mortality model predictions. Especially considering future climate-driven increases in fire frequency and intensity, knowledge of the physiological tree responses is important to better estimate postfire ecosystem dynamics and interactions with climate disturbances.

Deep neural networks (DNNs) for highly automated driving are often trained on a large and diverse dataset, and evaluation metrics are reported usually on a per-frame basis. However, when evaluated on video sequences, the predictions are often unstable between consecutive frames. As such unstable predictions over time can lead to severe safety consequences, there is a growing need to understand, evaluate, and improve the temporal consistency of DNNs. In this paper, we explore such a temporal characteristic and propose a novel unsupervised temporal consistency (TC) loss that penalizes unstable semantic segmentation predictions. This loss function is used in a two-stage training scheme to jointly optimize for both, accuracy of semantic segmentation predictions, and its temporal consistency based on video sequences. We demonstrate that our training strategy helps in improving the temporal consistency of two state-of-the-art semantic segmentation networks on two different road-scenes datasets. We report an absolute 4.25% improvement in the mean temporal consistency (mTC) of the HRNetV2 network and an absolute 2.78% improvement on the DeepLabv3+ network, both evaluated on the Cityscapes dataset, with only a slight decrease in accuracy. When evaluating on the same video sequences using a synthetic dataset Sim KI-A, we show absolute improvements in both, accuracy (2.19% mIoU) and temporal consistency (0.21% mTC) for the DeepLabv3+ network. We confirm similar improvements for the HRNetV2 network.

The heat resistance of meristematic tissues is crucial for the survival of plants exposed to high temperatures, as experienced during a forest fire. Although the risk and frequency of forest fires are increasing due to climate change, knowledge about the heat susceptibility of buds, which enclose apical meristems and thus enable resprouting and apical growth, is scarce. In this study, the heat resistance of buds in two different phenological stages was experimentally assessed for 10 European tree species. Cellular heat tolerance of buds was analyzed by determining the electrolyte leakage following heat exposure. Further, the heat insulation capability was tested by measuring the time required to reach lethal internal temperatures linked to bud traits. Our results highlighted differences in cellular heat tolerance and insulation capability among the study species. The phenological stage was found to affect both the thermal stability of cells and the buds' insulation. Further, a good relationship between size‐related bud traits and insulation capability was established. Species‐specific data on the heat resistance of buds give a more accurate picture of the fire susceptibility of European tree species and provide useful information for estimating tree post‐fire responses more precisely., The protection of meristematic tissues located in buds is crucial for the survival of plants exposed to high temperatures, as during a forest fire. This study demonstrated differences in cellular heat tolerance and insulation capability of buds among ten European tree species and evaluated the effect of the phenological stage on the buds' heat resistance.

Enabling autonomous driving (AD) can be considered one of the biggest challenges in today's technology. AD is a complex task accomplished by several functionalities, with environment perception being one of its core functions. Environment perception is usually performed by combining the semantic information captured by several sensors, i.e., lidar or camera. The semantic information from the respective sensor can be extracted by using convolutional neural networks (CNNs) for dense prediction. In the past, CNNs constantly showed state-of-the-art performance on several vision-related tasks, such as semantic segmentation of traffic scenes using nothing but the red-green-blue (RGB) images provided by a camera. Although CNNs obtain state-of-the-art performance on clean images, almost imperceptible changes to the input, referred to as adversarial perturbations, may lead to fatal deception. The goal of this article is to illuminate the vulnerability aspects of CNNs used for semantic segmentation with respect to adversarial attacks, and share insights into some of the existing known adversarial defense strategies. We aim to clarify the advantages and disadvantages associated with applying CNNs for environment perception in AD to serve as a motivation for future research in this field., Comment: IEEE Signal Processing Magazine (Volume: 38, Issue: 1, Jan. 2021), pp. 42 - 52