Explore the vast range of titles available.

Recent advances in wearable ultrasound technologies have demonstrated the potential for hands-free data acquisition, but technical barriers remain as these probes require wire connections, can lose track of moving targets and create data-interpretation challenges. Here we report a fully integrated autonomous wearable ultrasonic-system-on-patch (USoP). A miniaturized flexible control circuit is designed to interface with an ultrasound transducer array for signal pre-conditioning and wireless data communication. Machine learning is used to track moving tissue targets and assist the data interpretation. We demonstrate that the USoP allows continuous tracking of physiological signals from tissues as deep as 164 mm. On mobile subjects, the USoP can continuously monitor physiological signals, including central blood pressure, heart rate and cardiac output, for as long as 12 h. This result enables continuous autonomous surveillance of deep tissue signals toward the internet-of-medical-things. A wearable ultrasound patch monitors subjects in motion using machine learning and wireless electronics.

Image hash codes are produced by binarizing the embeddings of convolutional neural networks (CNN) trained for either classification or retrieval. While proxy embeddings achieve good performance on both tasks, they are non-trivial to binarize, due to a rotational ambiguity that encourages non-binary embeddings. The use of a fixed set of proxies (weights of the CNN classification layer) is proposed to eliminate this ambiguity, and a procedure to design proxy sets that are nearly optimal for both classification and hashing is introduced. The resulting hash-consistent large margin (HCLM) proxies are shown to encourage saturation of hashing units, thus guaranteeing a small binarization error, while producing highly discriminative hash-codes. A semantic extension (sHCLM), aimed to improve hashing performance in a transfer scenario, is also proposed. Extensive experiments show that sHCLM embeddings achieve significant improvements over state-of-the-art hashing procedures on several small and large datasets, both within and beyond the set of training classes.

Technological advances provide opportunities for automating direct observations of physical activity, which allow for continuous monitoring and feedback. This pilot study evaluated the initial validity of computer vision algorithms for ecological assessment of physical activity. The sample comprised 6630 seconds per camera (three cameras in total) of video capturing up to nine participants engaged in sitting, standing, walking, and jogging in an open outdoor space while wearing accelerometers. Computer vision algorithms were developed to assess the number and proportion of people in sedentary, light, moderate, and vigorous activity, and group-based metabolic equivalents of tasks (MET)-minutes. Means and standard deviations (SD) of bias/difference values, and intraclass correlation coefficients (ICC) assessed the criterion validity compared to accelerometry separately for each camera. The number and proportion of participants sedentary and in moderate-to-vigorous physical activity (MVPA) had small biases (within 20% of the criterion mean) and the ICCs were excellent (0.82–0.98). Total MET-minutes were slightly underestimated by 9.3–17.1% and the ICCs were good (0.68–0.79). The standard deviations of the bias estimates were moderate-to-large relative to the means. The computer vision algorithms appeared to have acceptable sample-level validity (i.e., across a sample of time intervals) and are promising for automated ecological assessment of activity in open outdoor settings, but further development and testing is needed before such tools can be used in a diverse range of settings.

To understand ocean health, it is crucial to monitor photosynthetic marine plankton – the microorganisms that form the base of the marine food web and are responsible for the uptake of atmospheric carbon. With the recent development of in situ microscopes that can acquire vast numbers of images of these organisms, the use of deep learning methods to taxonomically identify them has come to the forefront. Given this, two questions arise: 1) How well do deep learning methods such as Convolutional Neural Networks (CNNs) identify these marine organisms using data from in situ microscopes? 2) How well do CNN-derived estimates of abundance agree with established net and bottle-based sampling? Here, using images collected by the in situ Scripps Plankton Camera (SPC) system, we trained a CNN to recognize 9 species of phytoplankton, some of which are associated with Harmful Algal Blooms (HABs). The CNNs evaluated on 26 independent natural samples collected at Scripps Pier achieved an averaged accuracy of 92%, with 7 of 10 target categories above 85%. To compare abundance estimates, we fit a linear model between the number of organisms of each species counted in a known volume in the lab, with the number of organisms collected by the in situ microscope sampling at the same time. The linear fit between lab and in situ counts of several of the most abundant key HAB species suggests that, in the case of dinoflagellates, there is good correspondence between the two methods. As one advantage of our method, given the excellent correlation between lab counts and in situ microscope counts for key species, the methodology proposed here provides a way to estimate an equivalent volume in which the employed microscope can identify in-focus organisms and obtain statistically robust estimates of abundance.

Klebsiella pneumoniae has been singled out as an urgent threat to human health due to the increasing isolation of strains resistant to “last-line” antimicrobials, narrowing the treatment options against Klebsiella infections. Unfortunately, at present, we cannot identify candidate compounds in late-stage development for treatment of multidrug-resistant Klebsiella infections; this pathogen is exemplary of the mismatch between unmet medical needs and the current antimicrobial research and development pipeline. Furthermore, there is still limited evidence on K. pneumoniae pathogenesis at the molecular and cellular levels in the context of the interactions between bacterial pathogens and their hosts. In this research, we have uncovered a sophisticated strategy employed by Klebsiella to subvert the activation of immune defenses by controlling the modification of proteins. Our research may open opportunities to develop new therapeutics based on counteracting this Klebsiella- controlled immune evasion strategy. Klebsiella pneumoniae is an important cause of multidrug-resistant infections worldwide. Understanding the virulence mechanisms of K. pneumoniae is a priority and timely to design new therapeutics. Here, we demonstrate that K. pneumoniae limits the SUMOylation of host proteins in epithelial cells and macrophages (mouse and human) to subvert cell innate immunity. Mechanistically, in lung epithelial cells, Klebsiella increases the levels of the deSUMOylase SENP2 in the cytosol by affecting its K48 ubiquitylation and its subsequent degradation by the ubiquitin proteasome. This is dependent on Klebsiella preventing the NEDDylation of the Cullin-1 subunit of the ubiquitin ligase complex E3-SCF-βTrCP by exploiting the CSN5 deNEDDylase. Klebsiella induces the expression of CSN5 in an epidermal growth factor receptor (EGFR)-phosphatidylinositol 3-kinase (PI3K)-protein kinase B (AKT)-extracellular signal-regulated kinase (ERK)-glycogen synthase kinase 3 beta (GSK3β) signaling pathway-dependent manner. In macrophages, Toll-like receptor 4 (TLR4)-TRAM-TRIF-induced type I interferon (IFN) via IFN receptor 1 (IFNAR1)-controlled signaling mediates Klebsiella -triggered decrease in the levels of SUMOylation via let-7 microRNAs (miRNAs). Our results revealed the crucial role played by Klebsiella polysaccharides, the capsule, and the lipopolysaccharide (LPS) O-polysaccharide, to decrease the levels of SUMO-conjugated proteins in epithelial cells and macrophages. A Klebsiella -induced decrease in SUMOylation promotes infection by limiting the activation of inflammatory responses and increasing intracellular survival in macrophages. IMPORTANCE Klebsiella pneumoniae has been singled out as an urgent threat to human health due to the increasing isolation of strains resistant to “last-line” antimicrobials, narrowing the treatment options against Klebsiella infections. Unfortunately, at present, we cannot identify candidate compounds in late-stage development for treatment of multidrug-resistant Klebsiella infections; this pathogen is exemplary of the mismatch between unmet medical needs and the current antimicrobial research and development pipeline. Furthermore, there is still limited evidence on K. pneumoniae pathogenesis at the molecular and cellular levels in the context of the interactions between bacterial pathogens and their hosts. In this research, we have uncovered a sophisticated strategy employed by Klebsiella to subvert the activation of immune defenses by controlling the modification of proteins. Our research may open opportunities to develop new therapeutics based on counteracting this Klebsiella- controlled immune evasion strategy.

The problem of modeling the dynamic structure of human activities is considered. Video is mapped to a semantic feature space, which encodes activity attribute probabilities over time. The binary dynamic system (BDS) model is proposed to jointly learn the distribution and dynamics of activities in this space. This is a non-linear dynamic system that combines binary observation variables and a hidden Gauss–Markov state process, extending both binary principal component analysis and the classical linear dynamic systems. A BDS learning algorithm, inspired by the popular dynamic texture, and a dissimilarity measure between BDSs, which generalizes the Binet–Cauchy kernel, are introduced. To enable the recognition of highly non-stationary activities, the BDS is embedded in a bag of words. An algorithm is introduced for learning a BDS codebook, enabling the use of the BDS as a visual word for attribute dynamics (WAD). Short-term video segments are then quantized with a WAD codebook, allowing the representation of video as a bag-of-words for attribute dynamics. Video sequences are finally encoded as vectors of locally aggregated descriptors, which summarize the first moments of video snippets on the BDS manifold. Experiments show that this representation achieves state-of-the-art performance on the tasks of complex activity recognition and event identification.

We propose an adaptive model for backgrounds containing significant stochastic motion (e.g. water). The new model is based on a generalization of the Stauffer–Grimson background model, where each mixture component is modeled as a dynamic texture. We derive an online K-means algorithm for updating the parameters using a set of sufficient statistics of the model. Finally, we report on experimental results, which show that the proposed background model both quantitatively and qualitatively outperforms state-of-the-art methods in scenes containing significant background motions.

Machine learning methods have been widely employed for anomaly detection in time series data, but often struggle to identify rare anomalies in high-dimensional or non-stationary data. Gener- ative Adversarial Networks (GANs) have shown promise in addressing this limitation, but their effectiveness in detecting extremely rare anomalies remains a challenge. Additionally, the lack of systematic comparison methods for evaluating anomaly detection algorithms, particularly in relation to varying anomaly frequencies, has hindered progress in this field.This thesis addresses these challenges by introducing novel contributions to the realm of anomaly detection in time series data. Firstly, two new GAN-based architectures, TadGAN-DT and XTadGAN, are proposed to handle scenarios with extremely rare anomalies. The former, TadGAN-DT, incorporates non-parametric dynamic thresholding and pruning methods. The lat- ter, XTadGAN, leverages meta-information on expected anomaly frequencies to establish rarity- based dynamic thresholding and pruning strategies. Our experimental results demonstrate that both algorithms outperform other relevant approaches in rare anomaly detection.Furthermore, a comprehensive framework for evaluating anomaly detection models is intro- duced. This framework uses Monte Carlo sampling to generate an arbitrary number of time series from a small set of original datasets, simulating various controlled scenarios. It enables system- atic assessments across various time series attributes, specifically considering varying levels of anomaly rarity. This establishes a standardized test bench, facilitating a deeper understanding of model strengths and limitations. To enhance model comparisons, a novel sensitivity index, the x-score, is introduced. This metric provides an objective measure to evaluate the performance of different anomaly detection algorithms across a spectrum of attributes, particularly varying anomaly frequencies.This research contributes to the field of time series anomaly detection by advancing the under- standing of rare anomaly detection using GANs. It introduces a robust framework for systematic model evaluation, including a sensitivity index that enhances the reliability of model comparisons, guiding future research and improving the applicability of anomaly detection algorithms in real- world scenarios.

The benefits of integrating attention and object recognition are investigated. While attention is frequently modeled as a pre-processor for recognition, we investigate the hypothesis that attention is an intrinsic component of recognition and vice-versa. This hypothesis is tested with a recognition model, the hierarchical discriminant saliency network (HDSN), whose layers are top-down saliency detectors, tuned for a visual class according to the principles of discriminant saliency. As a model of neural computation, the HDSN has two possible implementations. In a biologically plausible implementation, all layers comply with the standard neurophysiological model of visual cortex, with sub-layers of simple and complex units that implement a combination of filtering, divisive normalization, pooling, and non-linearities. In a convolutional neural network implementation, all layers are convolutional and implement a combination of filtering, rectification, and pooling. The rectification is performed with a parametric extension of the now popular rectified linear units (ReLUs), whose parameters can be tuned for the detection of target object classes. This enables a number of functional enhancements over neural network models that lack a connection to saliency, including optimal feature denoising mechanisms for recognition, modulation of saliency responses by the discriminant power of the underlying features, and the ability to detect both feature presence and absence. In either implementation, each layer has a precise statistical interpretation, and all parameters are tuned by statistical learning. Each saliency detection layer learns more discriminant saliency templates than its predecessors and higher layers have larger pooling fields. This enables the HDSN to simultaneously achieve high selectivity to target object classes and invariance. The performance of the network in saliency and object recognition tasks is compared to those of models from the biological and computer vision literatures. This demonstrates benefits for all the functional enhancements of the HDSN, the class tuning inherent to discriminant saliency, and saliency layers based on templates of increasing target selectivity and invariance. Altogether, these experiments suggest that there are non-trivial benefits in integrating attention and recognition.