uuid
int64
0
6k
title
stringlengths
8
285
abstract
stringlengths
22
4.43k
100
ELECTRIcity: An Efficient Transformer for Non-Intrusive Load Monitoring
Non-Intrusive Load Monitoring (NILM) describes the process of inferring the consumption pattern of appliances by only having access to the aggregated household signal. Sequence-to-sequence deep learning models have been firmly established as state-of-the-art approaches for NILM, in an attempt to identify the pattern of the appliance power consumption signal into the aggregated power signal. Exceeding the limitations of recurrent models that have been widely used in sequential modeling, this paper proposes a transformer-based architecture for NILM. Our approach, called ELECTRIcity, utilizes transformer layers to accurately estimate the power signal of domestic appliances by relying entirely on attention mechanisms to extract global dependencies between the aggregate and the domestic appliance signals. Another additive value of the proposed model is that ELECTRIcity works with minimal dataset pre-processing and without requiring data balancing. Furthermore, ELECTRIcity introduces an efficient training routine compared to other traditional transformer-based architectures. According to this routine, ELECTRIcity splits model training into unsupervised pre-training and downstream task fine-tuning, which yields performance increases in both predictive accuracy and training time decrease. Experimental results indicate ELECTRIcity's superiority compared to several state-of-the-art methods.
101
Data-Efficient Deep Reinforcement Learning for Attitude Control of Fixed-Wing UAVs: Field Experiments
Attitude control of fixed-wing unmanned aerial vehicles (UAVs) is a difficult control problem in part due to uncertain nonlinear dynamics, actuator constraints, and coupled longitudinal and lateral motions. Current state-of-the-art autopilots are based on linear control and are thus limited in their effectiveness and performance. drl is a machine learning method to automatically discover optimal control laws through interaction with the controlled system that can handle complex nonlinear dynamics. We show in this article that deep reinforcement learning (DRL) can successfully learn to perform attitude control of a fixed-wing UAV operating directly on the original nonlinear dynamics, requiring as little as 3 min of flight data. We initially train our model in a simulation environment and then deploy the learned controller on the UAV in flight tests, demonstrating comparable performance to the state-of-the-art ArduPlane proportional-integral-derivative (PID) attitude controller with no further online learning required. Learning with significant actuation delay and diversified simulated dynamics were found to be crucial for successful transfer to control of the real UAV. In addition to a qualitative comparison with the ArduPlane autopilot, we present a quantitative assessment based on linear analysis to better understand the learning controller's behavior.
102
Sparse representation with multi-manifold analysis for texture classification from few training images
Texture classification is one of the most important tasks in computer vision field and it has been extensively investigated in the last several decades. Previous texture classification methods mainly used the template matching based methods such as Support Vector Machine and k-Nearest-Neighbour for classification. Given enough training images the state-of-the-art texture classification methods could achieve very high classification accuracies on some benchmark databases. However, when the number of training images is limited, which usually happens in real-world applications because of the high cost of obtaining labelled data, the classification accuracies of those state-of-the-art methods would deteriorate due to the overfitting effect. In this paper we aim to develop a novel framework that could correctly classify textural images with only a small number of training images. By taking into account the repetition and sparsity property of textures we propose a sparse representation based multi-manifold analysis framework for texture classification from few training images. A set of new training samples are generated from each training image by a scale and spatial pyramid, and then the training samples belonging to each class are modelled by a manifold based on sparse representation. We learn a dictionary of sparse representation and a projection matrix for each class and classify the test images based on the projected reconstruction errors. The framework provides a more compact model than the template matching based texture classification methods, and mitigates the overfitting effect. Experimental results show that the proposed method could achieve reasonably high generalization capability even with as few as 3 training images, and significantly outperforms the state-of-the-art texture classification approaches on three benchmark datasets. (C) 2014 Elsevier B.V. All rights reserved.
103
Dynamic graph convolutional network for multi-video summarization
Multi-video summarization is an effective tool for users to browse multiple videos. In this paper, multi-video summarization is formulated as a graph analysis problem and a dynamic graph convolutional network is proposed to measure the importance and relevance of each video shot in its own video as well as in the whole video collection. Two strategies are proposed to solve the inherent class imbalance problem of video summarization task. Moreover, we propose a diversity regularization to encourage the model to generate a diverse summary. Extensive experiments are conducted, and the comparisons are carried out with the state-of-the-art video summarization methods, the traditional and novel graph models. Our method achieves state-of-the-art performances on two standard video summarization datasets. The results demonstrate the effectiveness of our proposed model in generating a representative summary for multiple videos with good diversity. (C) 2020 Elsevier Ltd. All rights reserved.
104
Low temperature gradient thermoelectric generator: Modelling and experimental verification
Internet of Things (IoT) and wearable sensing paradigm assume the sensing devices are available 24/7 and can be accessed from anywhere. This vision implies strict requirements to the power supply and energy harvesting which are expected to guarantee 'perpetual' operation of IoT devices. This paper reports on modelling and experimental verification of low temperature gradient thermoelectric generator. Obtained under the conditions of low gradient temperature approximation, the model accounts for the key physical phenomena and enables the accurate output power calculations using a closed-form expression. We perform a comparative study on the state-of-the-art models against the obtained solution and show the simplicity and performance of the proposed approach. For demonstrating practical feasibility of the model, we develop an experimental testbed consisting of the power generator, temperature control and data acquisition units. Experimental results demonstrate the average error 5.5% which improves the state-of-the-art results.
105
A High-Accuracy and Energy-Efficient CORDIC Based Izhikevich Neuron With Error Suppression and Compensation
Bio-inspired neuron models are the key building blocks of brain-like neural networks for brain-science exploration and neuromorphic engineering applications. The efficient hardware design of bio-inspired neuron models is one of the challenges to implement brain-like neural networks, as the balancing of model accuracy, energy consumption and hardware cost is very challenging. This paper proposes a high-accuracy and energy-efficient Fast-Convergence COordinate Rotation DIgital Computer (FC-CORDIC) based Izhikevich neuron design. For ensuring the model accuracy, an error propagation model of the Izhikevich neuron is presented for systematic error analysis and effective error reduction. Parameter-Tuning Error Compensation (PTEC) method and Bitwidth-Extension Error Suppression (BEES) method are proposed to reduce the error of Izhikevich neuron design effectively. In addition, by utilizing the FC-CORDIC instead of conventional CORDIC for square calculation in the Izhikevich model, the redundant CORDIC iterations are removed and therefore, both the accumulated errors and required computation are effectively reduced, which significantly improve the accuracy and energy efficiency. An optimized fixed-point design of FC-CORDIC is also proposed to save hardware overhead while ensuring the accuracy. FPGA implementation results exhibit that the proposed Izhikevich neuron design can achieve high accuracy and energy efficiency with an acceptable hardware overhead, among the state-of-the-art designs.
106
High-Density Lipoprotein Cholesterol as a Potential Medium between Depletion of Lachnospiraceae Genera and Hypertension under a High-Calorie Diet
Gut microbial dysbiosis has been associated with hypertension. An extremely high incidence of essential hypertension was found in the Han and the Yugur people who resided in Sunan County in China's nomadic steppes, with little population movement. To investigate gut microbial contributions to this high incidence of hypertension, we recruited a total of 1, 242 Yugur and Han people, who had resided in Sunan County for more than 15 years and accounted for 3% of the local population. The epidemiological survey of 1,089 individuals indicated their nearly 1.8-times-higher prevalence of hypertension (38.2 to 43.3%) than the average in China (23.2%), under a special high-calorie diet based on wheat, cattle, mutton, and animal offal. Investigations of the fecal microbiota of another cohort of 153 individuals revealed that certain Lachnospiraceae genera were positively correlated with high-density lipoprotein cholesterol (HDL-C) but negatively correlated with systolic blood pressure (SBP) and diastolic blood pressure (DBP). HDL-C was negatively correlated with SBP and DBP. We further observed that the serum butyrate content was lower in both Han and Yugur people with hypertension than in those without hypertension. This study gives novel insight into the role of gut microbial dysbiosis in hypertension modulation under a high-calorie diet, where the notable depletion of Lachnospiraceae genera might lead to less production of butyrate, contributing to the lower level of HDL-C and elevating blood pressure in hypertension. IMPORTANCE Dietary nutrients can be converted by the gut microbiota into metabolites such as short-chain fatty acids, which may serve as disease-preventing agents in hypertension. Due to the limited population mobility and unique high-calorie dietary habits, the cohort of this study can serve as a representative cohort for elucidating the associations between the gut microbiota and hypertension under a high-calorie diet. Moreover, low levels of HDL-C have previously been associated with an increased risk of various cardiovascular diseases (CVDs). Our findings provide new insight showing that low levels of HDL-C may be a potential medium between the depletion of Lachnospiraceae genera and hypertension under a high-calorie diet, which might also be a potential candidate for other CVDs.
107
Novel Bacterial Topoisomerase Inhibitor Gepotidacin Demonstrates Absence of Fluoroquinolone-Like Arthropathy in Juvenile Rats
Fluoroquinolone use in children is limited due to its potential toxicity and negative effects on skeletal development, but the actual effects/risks of fluoroquinolones on bone growth and the mechanisms behind fluoroquinolone-driven arthropathy remain unknown. Gepotidacin is a novel, bactericidal, first-in-class triazaacenaphthylene antibiotic with a unique mechanism of action that is not anticipated to have the same risks to bone growth as those of fluoroquinolones. Gepotidacin is in phase III clinical development for uncomplicated urinary tract infections (ClinicalTrials.gov identifiers NCT04020341 and NCT04187144) and urogenital gonorrhea (ClinicalTrials.gov identifier NCT04010539) in adults and adolescents ≥12 years of age. To inform arthropathy and other potential toxicity risks of gepotidacin in pediatric studies, this nonclinical study assessed oral gepotidacin toxicity in juvenile rats from postnatal day (PND) 4 to PND 32/35 (approximately equivalent to human ages from newborn to 11 years), using both in-life assessments (tolerability, toxicity, and toxicokinetics) and terminal assessments (necropsy with macroscopic and microscopic skeletal femoral head and/or stifle joint examinations). Gepotidacin doses of ≤300 mg/kg of body weight/day were well tolerated from PND 4 to PND 21, and higher doses of ≤1,250 mg/kg/day were well tolerated from PND 22 when the dose levels were escalated to maintain systemic exposure levels up to PND 35, with no observed treatment-related clinical signs, effects on mean body weight gain, or macroscopic findings on articular surfaces. A dose of 1,000 mg/kg/day was not tolerated during the dosing period from PND 4 to 21, with effects on body weight gain, fecal consistency, and body condition. Microscopic effects on articular surfaces were evaluated after 32 days of gepotidacin treatment at the highest tolerated dose. After 32 days of treatment with the highest tolerated gepotidacin dose of 300/1,250 mg/kg/day (systemic concentrations [area under the curve {AUC} values] of 93.7 μg · h/mL [males] and 121 μg · h/mL [females]), no skeletal effects on articular surfaces of the femoral head or stifle joint were observed. The absence of treatment-related clinical signs and arthropathy in juvenile rats provides evidence to support the potential future use of gepotidacin in children.
108
Sinonasal Teratocarcinosarcoma Involving Orbital and Intracranial Extension: A Rare Case Report
Sinonasal teratocarcinosarcoma (SNTCS) is one of the rarest and most highly invasive malignant neoplasms often found in the nasal cavity and paranasal sinuses. SNTCS is often misdiagnosed because of its morphological heterogeneity. Due to its rarity, clinical characteristics and optimal therapy have not been well-established. Here, we present a case of SNTCS with orbital and intracranial extensions. A 48-year-old male patient presented with left-side nasal obstruction for 3 years. He appeared with visual and neurological symptoms 2 months ago. On radiographic examination, a mass was observed in the left paranasal sinuses with orbital and intracranial extension involvement. The mass was surgically resected. In the future, knowledge of this entity may assist in the accurate diagnosis and proper management of SNTCS.
109
Purinergic Signaling and Its Role in Mobilization of Bone Marrow Stem Cells
Mobilization or egress of stem cells from bone marrow (BM) into peripheral blood (PB) is an evolutionary preserved and important mechanism in an organism for self-defense and regeneration. BM-derived stem cells circulate always at steady-state conditions in PB, and their number increases during stress situations related to (a) infections, (b) tissue organ injury, (c) stress, and (d) strenuous exercise. Stem cells also show a circadian pattern of their PB circulating level with peak in early morning hours and nadir late at night. The number of circulating in PB stem cells could be pharmacologically increased after administration of some drugs such as cytokine granulocyte colony-stimulating factor (G-CSF) or small molecular antagonist of CXCR4 receptor AMD3100 (Plerixafor) that promote their egress from BM into PB and lymphatic vessels. Circulating can be isolated from PB for transplantation purposes by leukapheresis. This important homeostatic mechanism is governed by several intrinsic complementary pathways. In this chapter, we will discuss the role of purinergic signaling and extracellular nucleotides in regulating this process and review experimental strategies to study their involvement in mobilization of various types of stem cells that reside in murine BM.
110
Interleaved Sketch: Toward Consistent Network Telemetry for Commodity Programmable Switches
Network telemetry is vital to various network applications, including network anomaly detection, capacity planning, and congestion alleviation. State-of-the-art network telemetry systems are claimed to be scalable, flexible, all-purpose, and accurate. They adopt interval approaches that track network traffic in each interval and collect statistics for analysis at a specific epoch. However, interval methods are impaired by collecting inconsistency and clearing inconsistency, which pollute statistics. Moreover, The state-of-the-art centralized controllers have long latency, which aggravates the discrepancy. Accordingly, we propose the interleaved sketch, a consistent and decentralized network telemetry system across all switches. Each switch has two asymmetric sketches that work in an interleaved fashion, and is self-supervised to improve consistency. The distributed control plane extracts the flow characteristics and provides network-wide telemetry with low latency. We build a P4 prototype of our proposed interleaved sketch and test it on a Barefoot Tofino switch. Experimental results demonstrate that our interleaved sketch achieves ideal accuracy at line speed, with 6% resource overhead.
111
Ensemble-Based Out-of-Distribution Detection
To design an efficient deep learning model that can be used in the real-world, it is important to detect out-of-distribution (OOD) data well. Various studies have been conducted to solve the OOD problem. The current state-of-the-art approach uses a confidence score based on the Mahalanobis distance in a feature space. Although it outperformed the previous approaches, the results were sensitive to the quality of the trained model and the dataset complexity. Herein, we propose a novel OOD detection method that can train more efficient feature space for OOD detection. The proposed method uses an ensemble of the features trained using the softmax-based classifier and the network based on distance metric learning (DML). Through the complementary interaction of these two networks, the trained feature space has a more clumped distribution and can fit well on the Gaussian distribution by class. Therefore, OOD data can be efficiently detected by setting a threshold in the trained feature space. To evaluate the proposed method, we applied our method to various combinations of image datasets. The results show that the overall performance of the proposed approach is superior to those of other methods, including the state-of-the-art approach, on any combination of datasets.
112
Multiaxial Haar-Like Feature and Compact Cascaded Classifier for Versatile Recognition
A versatile recognition algorithm has been proposed to process image, sound, and 3-D acceleration signals with a common framework at low calculation cost. Firstly, a novel 1-D Haar-like feature is used to roughly extract frequency information from temporal signals. Biaxial and mean-embedded Haar-like features are proposed to extract the standard deviation and the interaxial correlation from 3-D acceleration signals. Secondly, two techniques are proposed to build a compact cascaded classifier. Redundant feature selection (RFS) incorporates the features which are already selected in previous stage classifiers to reduce the calculation cost. A dynamic look-up table (DLUT) is proposed to construct a look-up table-based weak classifier with the smallest possible number of bins. A train loss function is by globally optimized using dynamic programming. The proposed algorithm is tested experimentally on speech/nonspeech classification and human activity recognition. The proposed algorithm yields a speech/nonspeech classification performance comparable to the state-of-art method called MFCC while reducing the calculation cost by 100 times. The algorithm also achieves human activity recognition accuracy of 96.1% with calculation cost reduction of 84% compared with the state-of-art method based on C4.5 decision-tree classifier using the basic statistical features. The proposed algorithm has been employed to build the versatile recognition processor.
113
The effect of justified video game violence on aggressive behavior and moderated immersion: An experimental approach
The effect of violent video games on aggressive behavior is an important topic in the field of game research. Recently, growing evidence suggests that justified game violence decreases feelings of guilt caused by in-game immoral behavior. However, little is known about the impact on aggressive behavior, and whether other factors moderate this effect. In a two-factor experiment, we tested the impact of justification of video game violence on aggressive behavior, and whether this effect would be enhanced by game immersion. Pilot experiment 1 (N = 60) and pilot experiment 2 (N = 40) demonstrated that the justification of violence and game immersion was successfully controlled by avatar and graphics quality. In the Main experiment, 123 participants played one of four conditions of a video game (2 [justification: justified vs. unjustified violence] × 2 [immersion: high vs. low immersion]) and it was found that participants who played in the justified violence condition reported greater aggressive behavior than those in the unjustified violence condition. In addition, participants who played in high immersion reported greater aggressive behavior than those in low immersion. However, game immersion did not moderate the effects of justified violence. This unexpected effect is likely due to participants' distancing themselves from and identifying less with their violent avatars.
114
Definitions Matter: Dynamic Policy Framing of the Arts in Boston's Sustainable Cultural Development
Cultural sustainability has become a fourth pillar in sustainable development studies. Different from the research approach to embedding culture into conventional sustainable discourse, this article argues that the sustainability and resilience issues within the arts and cultural sector should be paid more attention to. Putting the arts and cultural sector in urban settings, sustainable cultural development entails dynamic policy framing and changing policy justifications in response to an evolving socioeconomic and political environment. Taking the policy framing of the arts as an analytical lens, this paper aims to investigate this dynamic change and key driving factors through an in-depth case study of Boston's urban cultural development. This article finds that different definitions of the arts are associated with different arts-based urban development strategies across four stages of cultural development in Boston spanning a period of over 75 years. The working definition moved from art to the arts, then to the creative arts industry, and eventually to cultural assets and creative capital. The policy framing of the arts keeps evolving and layering in pursuit of more legitimacy and resources regarding groups of stakeholders, field industry components, types of industrial structure, and multiple policy goals. This dynamic policy framing has been driven by arts advocacy groups, policy learning process, urban leadership change, and cultural institutional change, allowing Boston to draw on a growing and diversifying set of cultural resources in pursuit of sustainable cultural development.
115
Out of sight, out of mind: a distance-aware forgetting strategy for adaptive random testing
Adaptive random testing (ART) achieves better failure-detection effectiveness than random testing by increasing the diversity of test cases. However, the intention of ensuring even spread of test cases inevitably causes an overhead problem. Although two basic forgetting strategies (i.e. random forgetting and consecutive retention) were proposed to reduce the computation cost of ART, they only considered the temporal distribution of test cases. In the paper, we presented a distance-aware forgetting strategy for the fixed size candidate set version of ART (DF-FSCS), in which the spatial distribution of test cases is taken into consideration. For a given candidate, the test cases out of its sight are ignored to reduce the distance computation cost. At the same time, the dynamic adjustment for partitioning and the second-round forgetting are adopted to ensure the linear complexity of DF-FSCS algorithm. Both simulation analysis and empirical study are employed to investigate the efficiency and effectiveness of DF-FSCS. The experimental results show that DF-FSCS significantly outperforms the classical ART algorithm FSCS-ART in efficiency, and has comparable failure-detection effectiveness. Com-pared with two basic forgetting methods, DF-FSCS is better in both efficiency and effectiveness. In contrast with a typical linear-time ART algorithm RBCVT-Fast, our algorithm requires less computational overhead and exhibits the similar failure-detection capability. In addition, DF-FSCS has more reliable performance than RBCVT-Fast in detecting failures for the programs with high-dimensional input domain.
116
Decorated boulders and other neglected features of the Central Saharan rock art
Although the Central Saharan rock art has been studied for more than five decades, attention has been given almost exclusively to figurative paintings and engravings. Images were divided into stylistic groups, which functioned as a sequence of relative chronology. Non-figurative engravings, such as grooves, ovals, kettles and cupules, have been considered only marginally or were completely ignored. Next to the impressive images of animals and human beings, these small engravings were not even considered a true form of rock art and were never included into the stylistic rock art table. Even more surprising is the exclusion of Kel Essuf. These particular anthropomorphic petroglyphs were already known in the 1960s, however, because of their marked stylistic and thematic difference from other engravings, they were treated as anomalous and, like the non-figurative art, they were not included among the official styles. This paper focuses on previously neglected or unpublished rock art, and suggests that the simplest engravings are signs bearing their own meaning. When studied as a whole instead of being treated as isolated units without any relation to other forms of rock art, it emerges that grooves, ovals, cupules and Kel Essuf engravings were probably created following the same pattern.
117
Efficient Joint Estimation of Tracer Distribution and Background Signals in Magnetic Particle Imaging Using a Dictionary Approach
Background signals are a primary source of artifacts in magnetic particle imaging and limit the sensitivity of the method since background signals are often not precisely known and vary over time. The state-of-the art method for handling background signals uses one or several background calibration measurements with an empty scanner bore and subtracts a linear combination of these background measurements from the actual particle measurement. This approach yields satisfying results in case that the background measurements are taken in close proximity to the particle measurement and when the background signal drifts linearly. In this work, we propose a joint estimation of particle distribution and background signal based on a dictionary that is capable of representing typical background signals. Reconstruction is performed frame-by-frame with minimal assumptions on the temporal evolution of background signals. Thus, even non-linear temporal evolution of the latter can be captured. Using a singular-value decomposition, the dictionary is derived from a large number of background calibration scans that do not need to be recorded in close proximity to the particle measurement. The dictionary is sufficiently expressive and represented by its principle components. The proposed joint estimation of particle distribution and background signal is expressed as a linear Tikhonov-regularized least squares problem, which can be efficiently solved. In phantom experiments it is shown that the method strongly suppresses background artifacts and even allows to estimate and remove the direct feed-through of the excitation field.
118
Symmetry-aware graph matching
Visual symmetry encodes middle- to high-level geometric information and plays an important role in many computer vision applications. Not much effort, however, has been devoted to utilize symmetry information for graph matching. In this paper, we propose a new framework for symmetry-aware graph matching in the context of image matching. In the framework, we first define symmetry matrices to characterize the geometric symmetry relation among image features, and then develop two methods to discover such symmetry. After that we design general strategies to boost graph matching algorithms with symmetry constraints, and apply these strategies to several state-of-the-art algorithms. For evaluation, the proposed symmetry-aware graph matching algorithms are first applied to two datasets involving clear symmetry patterns: a recently proposed architecture image set and a dental panoramic radiograph collection. The results clearly demonstrate the benefits of using symmetry information. An experiment is also conducted on a dataset where symmetry is less obvious, the results show that our strategies still achieve better of similar performance compared with state-of-the-arts. (C) 2016 Elsevier Ltd. All rights reserved.
119
Archer: A History-Based Global Routing Algorithm
Global routing is an important step in the physical design process. In this paper, we propose a new global routing algorithm Archer, which resolves some of the most common problems with the state-of-the-art global routers. It is known that concurrent global routing algorithms are typically too expensive to be applied on today's large designs, which may contain up to a million nets. On the other hand, iterative rip-up and reroute (RNR)-based algorithms are susceptible to getting stuck in local optimal solutions. In this paper, we propose an RNR-based global routing algorithm that guides the routing iterations out of local optima through effective usage of congestion histories. We also focus on the problem of how to enable a smooth tradeoff between seemingly conflicting objectives of overflow and wirelength minimization. Furthermore, we propose a Lagrangian relaxation-based bounded-length min-cost topology improvement algorithm that enables Steiner trees to change dynamically for the purpose of congestion optimization. Our experiments on public benchmarks show the effectiveness of Archer compared to other state-of-the-art global routers.
120
Adaptation planning in France: Inputs from narratives of change in support of a community-led foresight process
In France, integrating adaptation to climate change into planning policies is a prerogative that has recently been delegated to municipalities. There are also various injunctions to engage the local population in this decision-making process. How can municipalities co-construct an adaptive future with their citizens? This article critically describes a community-led foresight process, based on the mapping, analysis and interpretation of narratives of change. Based on empirical results, we explore and discuss the role past, present and future narratives may play in the process of outlining incremental scenarios and how these might enable the identification of pathways and hinge points. The role of design in supporting the process by proposing an innovative foresight workshop is also discussed. We then highlight how these narratives stimulated reflections through an art, design and science foresight experiment.
121
Participatory arts and affective engagement with climate change: The missing link in achieving climate compatible behaviour change?
Despite a growing number of arts based climate change interventions and the importance emphasised in the social psychology literature of achieving affective (emotional) engagement with climate change before climate compatible behaviour change is likely, to date there has been no systematic application of interpretive social science techniques to understand the ways in which these arts based interventions do or do not achieve affective public engagement with climate change and hence might hold the key to unlocking broader climate compatible behaviour change. This article makes two key contributions.. First, it analyses the literature across social psychology and participatory arts to demonstrate why participatory, climate change based arts interventions could hold the key to more effective approaches to engaging multiple publics in climate compatible behaviour change. Second, using a small sample of participants in an arts based climate change intervention, in the Inner Hebrides, Scotland, it demonstrates the potential value of combining social science techniques (in this case Q Methodology) with participatory arts interventions to better understand and learn from the ways in which climate based arts interventions achieve affective public engagement with climate change. This raises the potential for, a significant new research and policy agenda looking forward.
122
Clinical Characteristics and Prognosis of Multiple Myeloma With Myelomatous Pleural Effusion: A Retrospective Single-Center Study
Objectives: Myelomatous pleural effusion is a rare presentation of extramedullary disease in multiple myeloma, which has been reported with dismal prognosis. We aimed to explore whether it has distinctive clinical characteristics and outcomes compared to other anatomic locations of extramedullary involvements. Methods: Multiple myeloma patients diagnosed at our institution from 2010 to 2020 were retrieved retrospectively. In total, 42 pairs of patients with and without extramedullary disease were enrolled, including 13 with myelomatous pleural effusion. The clinical and laboratory parameters were collected and compared between different groups. Prognostic effect of myelomatous pleural effusion was assessed in cox regression model and Kaplan-Meier curves. Results: Myelomatous pleural effusion patients presented a higher level of β2-microglobulin (P = .041), greater prevalence of multisites extramedullary lesions (69.2% vs 38.0%, P = .036) and International Staging System stage III (76.9% vs 44.8%, P = .016). Median overall survival was 60.6 months in patients without extramedullary disease versus 35.0 months in patients with extramedullary disease (P = .045). Notably, median overall survival was 13.0 months in myelomatous pleural effusion patients versus 37.0 months in other extramedullary disease patients with a significant difference (P = .029). Furtherly, multivariate analysis recognized myelomatous pleural effusion as an independent prognostic indicator (Hazard ratio: 2.669, 95% CI [1.132-6.293], P = .025). Conclusion: Myelomatous pleural effusion patients presented heavier tumor burden and worse outcomes than other extramedullary diseases.
123
New generation material for oil spill cleanup
Three-dimensional graphene-based materials promise oil spill cleanup in water at throughputs much higher than state-of-the-art oil-water separation materials.
124
Multi-Attribute Crowdsourcing Task Assignment With Stability and Satisfactory
Recently, crowdsourcing applications for smart cities have become more and more popular due to its higher work efficiency and lower work costs. However, the reasonable task assignment is still one of the important challenges for crowdsourcing. The existing researches on crowdsourcing task assignment focus on the tradeoff between maximizing the utility of platforms and minimizing the cost of requesters, but they lack of the considerations of stability and satisfactory. In this paper, we propose an intelligent multi attributes crowdsourcing task assignment with stability and satisfactory, called TASS. TASS can exploit the multi attributes to solve the stability of the transaction, and adopt the game theory to maximize the satisfaction of both sides during the task assignment. Next, we theoretically prove that the task assignment mechanism is truthfulness, individual rationality, stable and satisfactory assignment, and budget-balanced. Finally, we evaluate the performances of TASS with the state-of-the-art task assignment works. The experimental results show that TASS is better than the state-of-the-art task assignment works in terms of truthfulness, individually rationality, stable and satisfactory assignment, and balanced budget.
125
A novel fluorescence biosensor based on double-stranded DNA branch migration-induced HCR and DNAzyme feedback circuit for sensitive detection of Pseudomonas aeruginosa (clean version)
Pseudomonas aeruginosa (P. aeruginosa) is one of the most common bacteria in nosocomial infection. Here, a novel fluorescence biosensor based on double-stranded DNA branch migration-induced hybridization chain reaction (HCR) and DNAzyme feedback circuit was constructed for sensitive detection of P. aeruginosa. The binding of P. aeruginosa with its aptamer on a DNA three-way junction structure initiated the double-stranded DNA branch migration to form two DNA "Y" junction structures. One DNA "Y" junction structure opened the fluorescence-labelled DNA hairpins and triggered the HCR. The other DNA "Y" junction structure formed a double-stranded DNAzyme and cleaved the specific ribonucleotide site, producing new triggering probes to start the next cycle of the double-stranded DNA branch migration. Ultimately, a large number of DNA "Y" junction structures were produced, which greatly promoted signal amplification. Under optimized conditions, the proposed biosensor detected a wide linearity range of 102-107 CFU mL-1, and the limit of detection was 37 CFU mL-1 (S/N = 3). The recovery test results indicated that the biosensor has promising clinical application potential. Because of the simultaneous initiation of the HCR and the DNAzyme feedback circuit through the double-stranded DNA branch migration, the constructed biosensor provided an ideal platform for pathogenic bacteria detection without protein enzymes and complex signal amplification procedures.
126
Exploring variation in quality of care and clinical outcomes between neonatal units: a novel use for the UK National Neonatal Audit Programme (NNAP)
Neonatology is a relatively new specialty, in which much of the practice remains non-evidence based. Variation in the quality of care delivered is recognised but measuring this is challenging. One possible indicator of this is variation in practice. For more than a decade, the National Neonatal Audit Project (NNAP) has described variation in practice between UK neonatal units in relation to its annually reviewed audit measures. These are based on evidence based national standards or developed by a consensus method and have become de facto measures defining good quality of neonatal healthcare within the UK. In this article we explore the practicality of using the NNAP to look for associations between quality of care and outcomes. This would not be to validate the measures but could help towards a better understanding of the reasons underlying recognised variation in outcomes, even between neonatal units of the same designation.
127
A Rare Case of a Vocal Cord Foreign Body in an Infant: A Case Report
A foreign body (FB) is an object or item that is foreign to the area in which it is found. FB in the airway, accompanied by the esophagus, is a common overnight emergency in pediatric otolaryngology. Here we report a case of a healthy 11-month-old girl who presented in the emergency room with stridor and a weak cry. The patient was admitted as a case of croup (laryngotracheobronchitis) and treated with multiple antibiotics for more than five days but showed no improvement, then consulted the ear, nose, and throat team (ENT).
128
Overview of efficient high-quality state-of-the-art depth enhancement methods by thorough design space exploration
High-quality 3D content generation requires high-quality depth maps. In practice, depth maps generated by stereo-matching, depth sensing cameras, or decoders, have low resolution and suffer from unreliable estimates and noise. Therefore, depth enhancement is necessary. Depth enhancement comprises two stages: depth upsampling and temporal post-processing. In this paper, we extend our previous work on depth upsampling in two ways. First we propose PWAS-MCM, a new depth upsampling method, and we show that it achieves on average the highest depth accuracy compared to other efficient state-of-the-art depth upsampling methods. Then, we benchmark all relevant state-of-the-art filter-based temporal post-processing methods on depth accuracy by conducting a parameter space search to find the optimum set of parameters for various upscale factors and noise levels. Then we analyze the temporal post-processing methods qualitatively. Finally, we analyze the computational complexity of each depth upsampling and temporal post-processing method by measuring the throughput and hardware utilization of the GPU implementation that we built for each method.
129
Novel metal chelating molecules with anticancer activity. Striking effect of the imidazole substitution of the histidine-pyridine-histidine system
Previously we have reported a metal chelating histidine-pyridine-histidine system possessing a trityl group on the histidine imidazole, namely HPH-2Trt, which induces apoptosis in human pancreatic adenocarcinoma AsPC-1 cells. Herein the influence of the imidazole substitution of HPH-2Trt was examined. Five related compounds, HPH-1Trt, HPH-2Bzl, HPH-1Bzl, HPH-2Me, and HPH-1Me were newly synthesized and screened for their activity against AsPC-1 and brain tumor cells U87 and U251. HPH-1Trt and HPH-2Trt were highly active among the tested HPH compounds. In vitro DNA cleavage assay showed both HPH-1Trt and HPH-2Trt completely disintegrate pUC19 DNA. The introduction of trityl group decisively potentiated the activity.
130
A local search method based on edge age strategy for minimum vertex cover problem in massive graphs
Minimum vertex cover problem (MVC) is a classic combinatorial optimization problem, which has many critical real-life applications in scheduling, VLSI design, artificial intelligence, and network security. For MVC, researchers have proposed many heuristic algorithms, especially local search algorithms. And recently, researchers have increased their interest in solving large real-world graphs which require algorithms with faster searching performance. In this work, we propose a new edge weighting method called EABMS. EABMS has a time complexity of O(1). Based on EABMS, we propose our MVC solver framework called EAVC in solving MVC for massive graphs. We conducted experiments and compared the results of EAVC solvers with state of the art solvers. The results show that EABMS is effective in weighing edges for large sparse graphs and EAVC solvers outperform state of the art solvers.
131
Generalizing state-of-the-art object detectors for autonomous vehicles in unseen environments
In scene understanding for autonomous vehicles (AVs), models trained on the available datasets fail to generalize well to the complex, real-world scenarios with higher dynamics. In this work, we attempt to handle the distribution mismatch by employing the generative adversarial network (GAN) and weather modeling to strengthen the intra-domain data. We also alleviate the fragility of our trained models against natural distortions with stateof-the-art augmentation approaches. Finally, we assess our method for cross-domain object detection through CARLA simulation. Our experiments demonstrate that: (1) Augmenting training class with even limited intradomain data captured from the adverse weather conditions boosts the generalization of the two kinds of object detectors; (2) Exploiting GANs and weather modeling to elaborately simulate the adverse, intra-domain weather conditions manages to surmount the adverse data scarcity issue for intra-domain object detection; (3) A combination of Augmix and style augmentations not only can promote the robustness of our trained models against different natural distortions but also can boost their performance in the cross-domain object detection; (4) Training GANs for unsupervised image-to-image translation by means of the existing, large-scale datasets outside of our training domain is found beneficial to alleviate image-based and instance-based domain shifts.
132
Efficient Minimum Flow Decomposition via Integer Linear Programming
Minimum flow decomposition (MFD) is an NP-hard problem asking to decompose a network flow into a minimum set of paths (together with associated weights). Variants of it are powerful models in multiassembly problems in Bioinformatics, such as RNA assembly. Owing to its hardness, practical multiassembly tools either use heuristics or solve simpler, polynomial time-solvable versions of the problem, which may yield solutions that are not minimal or do not perfectly decompose the flow. Here, we provide the first fast and exact solver for MFD on acyclic flow networks, based on Integer Linear Programming (ILP). Key to our approach is an encoding of all the exponentially many solution paths using only a quadratic number of variables. We also extend our ILP formulation to many practical variants, such as incorporating longer or paired-end reads, or minimizing flow errors. On both simulated and real-flow splicing graphs, our approach solves any instance in <13 seconds. We hope that our formulations can lie at the core of future practical RNA assembly tools. Our implementations are freely available on Github.
133
Sentence-Based Sentiment Analysis for Expressive Text-to-Speech
Current research to improve state of the art Text-To-Speech (TTS) synthesis studies both the processing of input text and the ability to render natural expressive speech. Focusing on the former as a front-end task in the production of synthetic speech, this article investigates the proper adaptation of a Sentiment Analysis procedure (positive/neutral/negative) that can then be used as an input feature for expressive speech synthesis. To this end, we evaluate different combinations of textual features and classifiers to determine the most appropriate adaptation procedure. The effectiveness of this scheme for Sentiment Analysis is evaluated using the Semeval 2007 dataset and a Twitter corpus, for their affective nature and their granularity at the sentence level, which is appropriate for an expressive TTS scenario. The experiments conducted validate the proposed procedure with respect to the state of the art for Sentiment Analysis.
134
Leaves of Persimmon (Diospyros kaki Thunb.) Ameliorate N-Methyl-N-nitrosourea (MNU)-Induced Retinal Degeneration in Mice
The purpose of the study was to investigate the protective effects of the ethanol extract of Diospyros kaki (EEDK) persimmon leaves to study N-methyl-N-nitrosourea (MNU)-induced retinal degeneration in mice. EEDK was orally administered after MNU injection. Retinal layer thicknesses were significantly increased in the EEDK-treated group compared with the MNU-treated group. The outer nuclear layer was preserved in the retinas of EEDK-treated mice. Moreover, EEDK treatment reduced the MNU-dependent up-regulation of glial fibrillary acidic protein (GFAP) and nestin expression in Müller and astrocyte cells. EEDK treatment also inhibited MNU-dependent down-regulation of rhodopsin expression. Quercetin exposure significantly attenuated the negative effects of H2O2 in R28 cells, suggesting that quercetin can act in an antioxidative capacity. Thus, EEDK may be considered as an agent for treating or preventing degenerative retinal diseases, such as retinitis pigmentosa and age-related macular degeneration.
135
Coarse-to-Fine Semantic Segmentation From Image-Level Labels
Deep neural network-based semantic segmentation generally requires large-scale cost extensive annotations for training to obtain better performance. To avoid pixel-wise segmentation annotations that are needed for most methods, recently some researchers attempted to use object-level labels (e.g., bounding boxes) or image-level labels (e.g., image categories). In this paper, we propose a novel recursive coarse-to-fine semantic segmentation framework based on only image-level category labels. For each image, an initial coarse mask is first generated by a convolutional neural network-based unsupervised foreground segmentation model and then is enhanced by a graph model. The enhanced coarse mask is fed to a fully convolutional neural network to be recursively refined. Unlike the existing image-level label-based semantic segmentation methods, which require labeling of all categories for images that contain multiple types of objects, our framework only needs one label for each image and can handle images that contain multi-category objects. Only trained on ImageNet, our framework achieves comparable performance on the PASCAL VOC dataset with other image-level label-based state-of-the-art methods of semantic segmentation. Furthermore, our framework can be easily extended to foreground object segmentation task and achieves comparable performance with the state-of-the-art supervised methods on the Internet object dataset.
136
Enhanced recovery after surgery in transurethral surgery for benign prostatic hyperplasia
Enhanced recovery after surgery (ERAS) measures have not been systematically applied in transurethral surgery for benign prostatic hyperplasia (BPH). This study was performed on patients with BPH who required surgical intervention. From July 2019 to June 2020, the ERAS program was applied to 248 patients, and the conventional program was applied to 238 patients. After 1 year of follow-up, the differences between the ERAS group and the conventional group were evaluated. The ERAS group had a shorter time of urinary catheterization compared with the conventional group (mean ± standard deviation [s.d.]: 1.0 ± 0.4 days vs 2.7 ± 0.8 days, P < 0.01), and the pain (mean ± s.d.) was significantly reduced through postoperative hospitalization days (PODs) 0-2 (POD 0: 1.7 ± 0.8 vs 2.4 ± 1.0, P < 0.01; POD 1: 1.6 ± 0.9 vs 3.5 ± 1.3, P < 0.01; POD 2: 1.2 ± 0.7 vs 3.0 ± 1.3, P < 0.01). No statistically significant difference was found in the rate of postoperative complications, such as postoperative bleeding (P = 0.79), urinary retention (P = 0.40), fever (P = 0.55), and readmission (P = 0.71). The hospitalization cost of the ERAS group was similar to that of the conventional group (mean ± s.d.: 16 927.8 ± 5808.1 Chinese Yuan [CNY] vs 17 044.1 ± 5830.7 CNY, P =0.85). The International Prostate Symptom Scores (IPSS) and quality of life (QoL) scores in the two groups were also similar when compared at 1 month, 3 months, 6 months, and 12 months after discharge. The ERAS program we conducted was safe, repeatable, and efficient. In conclusion, patients undergoing the ERAS program experienced less postoperative stress than those undergoing the conventional program.
137
Dynamic Interactions Between Brilliant Green and MscL Investigated by Solid-State NMR Spectroscopy and Molecular Dynamics Simulations
The mechanosensitive ion channel of large conductance (MscL) is a promising template for the development of new antibiotics due to its high conservation and uniqueness to microbes. Brilliant green (BG), a triarylmethane dye, has been identified as a new antibiotic targeted MscL. However, the detailed binding sites to MscL and the dynamic pathway of BG through the MscL channel remain unknown. Here, the dynamic interactions between BG and MscL were investigated using solid-state NMR spectroscopy and molecule dynamics (MD) simulations. Residue site-specific binding sites of BG to the MscL channel were identified by solid-state NMR. In addition, MD simulations revealed that BG conducts through the MscL channel via residues along the inner surface of the pore sequentially, in which the strong hydrophobic interactions between BG and hydrophobic residues F23 and I27 in the hydrophobic gate region of the MscL channel are major restrictions. Particularly, it was demonstrated that BG activates the MscL channel by reducing the hydrophobicity of the F23 in the gate region by water molecules that are bound to BG. Taken together, these simulations and experimental data provide novel insights into the dynamic interactions between BG and MscL, based on which new hydrophobic antibiotics and adjuvants targeting MscL can be developed.
138
A green and environmental sustainable approach to synthesis the Mn oxide nanomaterial from Punica granatum leaf extracts and its in vitro biological applications
Pathogenic fungal infections in fruit cause economic losses and have deleterious effects on human health globally. Despite the low pH and high water contents of vegetables and fresh, ripened fruits, they are prone to fungal and bacterial diseases. The ever-increasing resistance of phytopathogens toward pesticides, fungicides and bactericides has resulted in substantial threats to plant growth and production in recent years. However, plant-mediated nanoparticles are useful tools for combating parasitic fungi and bacteria. Herein, we synthesized biogenic manganese oxide nanoparticles (MnONPs) from an extract of Punica granatum (P. granatum), and these nanoparticles showed significant antifungal and antibacterial activities. The production of MnONPs from plant extracts was confirmed by infrared spectroscopy (FTIR), X-ray diffraction (XRD) and UV visible spectroscopy (UV). The surface morphology and shape of the nanoparticles were characterized by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Using a detached fruit method, the MnONPs were shown to exhibit significant antimicrobial activities against two bacterial strains, E. coli and S. aureus, and against the fungal species P. digitatum. The results revealed that the MnONPs had a minimum antimicrobial activity at 25 µg/mL and a maximum antimicrobial activity at 100 µg/mL against bacterial strains in lemon (citrus). Furthermore, the MnONPs exhibited significant ROS scavenging activity. Finally, inconclusive results from the green-synthesized MnONPs magnified their significant synergetic effects on the shelf life of tomatoes (Lycopercicum esculantum) and indicated that they could be used to counteract the phytopathological effects of postharvest fungal diseases in fruits and vegetables. Overall, this method of MnONPs synthesis is inexpensive, rapid and ecofriendly. MnONPs can be used as potential antimicrobial agents against different microbial species.
139
CANet: Context Aware Network for Brain Glioma Segmentation
Automated segmentation of brain glioma plays an active role in diagnosis decision, progression monitoring and surgery planning. Based on deep neural networks, previous studies have shown promising technologies for brain glioma segmentation. However, these approaches lack powerful strategies to incorporate contextual information of tumor cells and their surrounding, which has been proven as a fundamental cue to deal with local ambiguity. In this work, we propose a novel approach named Context-Aware Network (CANet) for brain glioma segmentation. CANet captures high dimensional and discriminative features with contexts from both the convolutional space and feature interaction graphs. We further propose context guided attentive conditional random fields which can selectively aggregate features. We evaluate our method using publicly accessible brain glioma segmentation datasets BRATS2017, BRATS2018 and BRATS2019. The experimental results show that the proposed algorithm has better or competitive performance against several State-of-The-Art approaches under different segmentation metrics on the training and validation sets.
140
Electrochemical Response of Redox Amino Acid Encoded Fluorescence Protein for Hydroxychloroquine Sensing
The sudden rise in the demand has led to large-scale production of hydroxychloroquine (HCQ) in the global market for various diseases such as malaria, rheumatic arthritis, and systemic lupus erythematous and prophylactic treatment of early SARS-CoV-2 outbreak. Thorough monitoring of HCQ intake patients is in high demand; hence, we have developed a redox amino acid encoded fluorescent protein-based electrochemical biosensor for sensitive and selective detection of HCQ. This electrochemical biosensor is generated based on the two-electron transfer process between redox amino acid (3,4-dihydroxy-L-phenylalanine, DOPA) encoded bio-redox protein and the HCQ forms the conjugate. The DOPA residue in the bio-redox protein specifically binds with HCQ, thereby producing a remarkable electrochemical response on the glassy carbon electrode. Experimental results show that the developed biosensor selectively and sensitively detects the HCQ in spiked urine samples. The reagent-free bio-redox capacitor detects HCQ in the range of 90 nM to 4.4 µM in a solution with a detection limit of 58 nM, signal to noise ratio of 3:1, and strong anti-interference ability. Real-time screening, quantification, and relative mean recoveries of HCQ on spiked urine samples were monitored through electron shuttling using bio-redox protein and were found to be 97 to 101%. Overall, the developed bio-redox protein-based sensor has specificity, selectivity, reproducibility, and sensitivity making it potentially attractive for the sensing of HCQ and also applicable to clinical research.
141
A comparison between state-of-the-art and neural network modelling of solar collectors
The state-of-the-art modelling of solar collectors as described in the European Standard EN 12975-2 is based on equations describing the thermal behaviour of the collectors by characterising the physical phenomena, e.g. transmission of irradiance through transparent covers, absorption of irradiance by the absorber, temperature dependent heat losses and others. This approach leads to so called collector parameters that describe these phenomena, e.g. the zero-loss collector efficiency eta(0) or the heat loss coefficients a(1) and a(2). Although the state-of-the-art approach in collector modelling and testing fits most of the collector types very well there are some collector designs (e.g. "Sydney" tubes using heat pipes and "water-in-glass" collectors) which cannot be modelled with the same accuracy than conventional collectors like flat plate or standard evacuated tubular collectors. The artificial neural network (ANN) approach could be an appropriate alternative to overcome this drawback. To compare the different approaches of modelling investigations for a conventional flat plate collector and an evacuated "Sydney" tubular collector have been carried out based on performance measurements according to the European Standard EN 12975-2. The investigations include the parameter identification (training), the comparisons between measured and modelled collector output and the simulated yearly collector yield for a solar domestic hot water system for both models. The obtained results show better agreement between measured and calculated collector output for the artificial neural network approach compared with the state-of-the-art modelling. The investigations also show that for the ANN approach special test sequences have to be designed and that the determination of the ANN that fits the thermal performance of the collector in the best way depends significantly on the expertise of the user. Nevertheless artificial neural networks have the potential to become an interesting alternative to the state-of-the-art collector models used today. (C) 2012 Elsevier Ltd. All rights reserved.
142
Automating design with intelligent human-machine integration
This paper reviews the state-of-the-art methodologies for automating design with intelligent human-machine integration from the perspectives of ontology and epistemology. The human-machine integrated automating design paradigm is reviewed systematically based on a proposed prototype of human-machine integrated design, from the aspects of ontology-based knowledge management with local-to-global ontology transitions, and epistemology-based upward-spiral cognitive process of coupled design ideation. Particularly, imaginal thinking frame is proposed as the foundation of intelligent human-machine interaction that puts human and machine on an equal platform. Further, this paper presents implementations and applications of the automating design paradigm and concludes with the identification of future trend. (C) 2015 CIRP.
143
Gradients of Orientation, Composition, and Hydration of Proteins for Efficient Light Collection by the Cornea of the Horseshoe Crab
The lateral eyes of the horseshoe crab, Limulus polyphemus, are the largest compound eyes within recent Arthropoda. The cornea of these eyes contains hundreds of inward projecting elongated cuticular cones and concentrate light onto proximal photoreceptor cells. Although this visual system has been extensively studied before, the precise mechanism allowing vision has remained controversial. Correlating high-resolution quantitative refractive index (RI) mapping and structural analysis, it is demonstrated how gradients of RI in the cornea stem from structural and compositional gradients in the cornea. In particular, these RI variations result from the chitin-protein fibers architecture, heterogeneity in protein composition, and bromine doping, as well as spatial variation in water content resulting from matrix cross-linking on the one hand and cuticle porosity on the other hand. Combining the realistic cornea structure and measured RI gradients with full-wave optical modeling and ray tracing, it is revealed that the light collection mechanism switches from refraction-based graded index (GRIN) optics at normal light incidence to combined GRIN and total internal reflection mechanism at high incident angles. The optical properties of the cornea are governed by different mechanisms at different hierarchical levels, demonstrating the remarkable versatility of arthropod cuticle.
144
An Efficient Eulerian Video Magnification Technique for Micro-biology Applications
The micro-biology videos often contain motions of particles which are not visible to naked eye. Therefore an efficient motion magnification technique is required to magnify these motions. A time efficient Eulerian video magnification technique for micro-biological applications is proposed. The proposed technique utilizes the concept of time and spatial uniformity to reduce the computational complexity. Simulation results reveal that the proposed scheme is almost four times efficient and more accurate as compared to state of art video magnification technique.
145
Improved Highway Network Block for Training Very Deep Neural Networks
Very deep networks are successful in various tasks with reported results surpassing human performance. However, training such very deep networks is not trivial. Typically, the problems of learning the identity function and feature reuse can work together to plague optimization of very deep networks. In this paper, we propose a highway network with gate constraints that addresses the aforementioned problems, and thus alleviates the difficulty of training. Namely, we propose two variants of highway network, HWGC and HWCC, employing feature summation and concatenation respectively. The proposed highway networks, besides being more computationally efficient, are shown to have more interesting learning characteristics such as natural learning of hierarchical and robust representations due to a more effective usage of model depth, fewer gates for successful learning, better generalization capacity and faster convergence than the original highway network. Experimental results show that our models outperform the original highway network and many state-of-the-art models. Importantly, we observe that our second model with feature concatenation and compression consistently outperforms our model with feature summation of similar depth, the original highway network, many state-of-the-art models and even ResNets on four benchmarking datasets which are CIFAR-10, CIFAR-100, Fashion-MNIST, SVHN and imagenet-2012 (ILSVRC) datasets. Furthermore, the second proposed model is more computationally efficient than the state-of-the-art in view of training, inference time and GPU memory resource, which strongly supports real-time applications. Using a similar number of model parameters for the CIFAR-10, CIFAR-100, Fashion-MNIST and SVHN datasets, the significantly shallower proposed model can surpass the performance of ResNet-110 and ResNet-164 that are roughly 6 and 8 times deeper, respectively. Similarly, for the imagenet dataset, the proposed models surpass the performance of ResNet-101 and ResNet-152 that are roughly three times deeper.
146
A chemical timer approach to delayed chemiluminescence
Although the onset time of chemical reactions can be manipulated by mechanical, electrical, and optical methods, its chemical control remains highly challenging. Herein, we report a chemical timer approach for manipulating the emission onset time of chemiluminescence (CL) reactions. A mixture of Mn2+, NaHCO3, and a luminol analog with H2O2 produced reactive oxygen species (ROS) radicals and other superoxo species (superoxide containing complex) with high efficiency, accompanied by strong and immediate CL emission. Surprisingly, the addition of thiourea postponed CL emission in a concentration-dependent manner. The delay was attributed to a slow-generation-scavenging mechanism, which was found to be generally applicable not only to various types of CL reagents and ROS radical scavengers but also to popular chromogenic reactions. The precise regulation of CL kinetics was further utilized in dynamic chemical coding with improved coding density and security. This approach provides a powerful platform for engineering chemical reaction kinetics using chemical timers, which is of application potential in bioassays, biosensors, CL microscopic imaging, microchips, array chips, and informatics.
147
YOLO-RTUAV: Towards Real-Time Vehicle Detection through Aerial Images with Low-Cost Edge Devices
Object detection in aerial images has been an active research area thanks to the vast availability of unmanned aerial vehicles (UAVs). Along with the increase of computational power, deep learning algorithms are commonly used for object detection tasks. However, aerial images have large variations, and the object sizes are usually small, rendering lower detection accuracy. Besides, real-time inferencing on low-cost edge devices remains an open-ended question. In this work, we explored the usage of state-of-the-art deep learning object detection on low-cost edge hardware. We propose YOLO-RTUAV, an improved version of YOLOv4-Tiny, as the solution. We benchmarked our proposed models with various state-of-the-art models on the VAID and COWC datasets. Our proposed model can achieve higher mean average precision (mAP) and frames per second (FPS) than other state-of-the-art tiny YOLO models, especially on a low-cost edge device such as the Jetson Nano 2 GB. It was observed that the Jetson Nano 2 GB can achieve up to 12.8 FPS with a model size of only 5.5 MB.
148
How Context or Knowledge Can Benefit Healthcare Question Answering?
Healthcare question answering (HQA) is a challenging task as questions are generally non-factoid in nature. Traditional information retrieval techniques do not perform well on non-factoid questions. Recent neural question answering systems are reported to have performance gains over traditional methods. However, little attention has been given to HQA as datasets are generally too small to train a neural model from scratch. Recently, several systems have been proposed to learn context representations for HQA. Despite moderate progress, these systems have not been thoroughly compared with state-of-the-art neural models, and these neural models were tested only on self-created datasets. This makes it difficult for practitioners to decide which models should be used for various scenarios. To address the above challenges, we develop a new joint model to incorporate both context and knowledge embeddings into neural ranking architectures. First, we adapt context embedding pre-trained from large open-domain corpus to small healthcare datasets. Second, we learn knowledge embedding from knowledge graphs to provide external information for understanding non-factoid questions. To evaluate how context or knowledge embedding can benefit HQA, we adapt many state-of-the-art methods for general QA to HQA, by injecting the context or knowledge information only, or both of them. Extensive experiments are conducted to compare our approach with those adapted methods and current HQA systems. The results show that our approach achieves the state-of-the-art performance on both HealthQA and NFCorpus datasets.
149
SfSNet: Learning Shape, Reflectance and Illuminance of Faces in the Wild
We present SfSNet, an end-to-end learning framework for producing an accurate decomposition of an unconstrained human face image into shape, reflectance and illuminance. SfSNet is designed to reflect a physical lambertian rendering model. SfSNet learns from a mixture of labeled synthetic and unlabeled real-world images. This allows the network to capture low-frequency variations from synthetic and high-frequency details from real images through the photometric reconstruction loss. SfSNet consists of a new decomposition architecture with residual blocks that learns a complete separation of albedo and normal. This is used along with the original image to predict lighting. SfSNet produces significantly better quantitative and qualitative results than state-of-the-art methods for inverse rendering and independent normal and illumination estimation. We also introduce a companion network, SfSMesh, that utilizes normals estimated by SfSNet to reconstruct a 3D face mesh. We demonstrate that SfSMesh produces face meshes with greater accuracy than state-of-the-art methods on real-world images.
150
Exploiting All Programmable SoCs in Neural Signal Analysis: A Closed-Loop Control for Large-Scale CMOS Multielectrode Arrays
Microelectrode array (MEA) systems with up to several thousands of recording electrodes and electrical or optical stimulation capabilities are commercially available or described in the literature. By exploiting their submillisecond and micrometric temporal and spatial resolutions to record bioelectrical signals, such emerging MEA systems are increasingly used in neuroscience to study the complex dynamics of neuronal networks and brain circuits. However, they typically lack the capability of implementing real-time feedback between the detection of neuronal spiking events and stimulation, thus restricting large-scale neural interfacing to open-loop conditions. In order to exploit the potential of such large-scale recording systems and stimulation, we designed and validated a fully reconfigurable FPGA-based processing system for closed-loop multichannel control. By adopting a Xilinx Zynq-all-programmable system on chip that integrates reconfigurable logic and a dual-core ARM-based processor on the same device, the proposed platform permits low-latency preprocessing (filtering and detection) of spikes acquired simultaneously from several thousands of electrode sites. To demonstrate the proposed platform, we tested its performances through ex vivo experiments on the mice retina using a state-of-the-art planar high-density MEA that samples 4096 electrodes at 18 kHz and record light-evoked spikes from several thousands of retinal ganglion cells simultaneously. Results demonstrate that the platform is able to provide a total latency from whole-array data acquisition to stimulus generation below 2 ms. This opens the opportunity to design closed-loop experiments on neural systems and biomedical applications using emerging generations of planar or implantable large-scale MEA systems.
151
Simultaneous Registration of Location and Orientation in Intravascular Ultrasound Pullbacks Pairs Via 3D Graph-Based Optimization
A novel method is reported for simultaneous registration of location (axial direction) and orientation (circumferential direction) of two intravascular ultrasound (IVUS) pullbacks of the same vessel taken at different times. Monitoring plaque progression or regression (e.g., during lipid treatment) is of high clinical relevance. Our method uses a 3D graph optimization approach, in which the cost function jointly reflects similarity of plaque morphology and plaque/perivascular image appearance. Graph arcs incorporate prior information about temporal correspondence of the two IVUS sequences and limited angular twisting between consecutive IVUS images. Additionally, our approach automatically identifies starting and ending frame pairs in the two IVUS pullbacks. Validation of our method was performed in 29 pairs of IVUS baseline/follow-up pullback sequences consisting of 8 622 IVUS image frames in total. In comparison to manual registration by three experts, the average location and orientation registration errors ranged from 0.72 mm to 0.79 mm and from 7.3 degrees to 9.3 degrees, respectively, all close to the inter-observer variability with no difference being statistically significant (p = NS) . Rotation angles determined by our automated approach and expert observers showed high correlation (r(2) of 0.97 to 0.98) and agreed closely (mutual bias between the automated method and expert observers ranged from -1.57 degrees to 0.15 degrees). Compared with state-of-the-art approaches, the new method offers lower errors in both location and orientation registration. Our method offers highly automated and accurate IVUS pullback registration and can be employed in IVUS-based studies of coronary disease progression, enabling more focal studies of coronary plaque development and transition of vulnerability.
152
Phenotypic variability in 446 CADASIL patients: Impact of NOTCH3 gene mutation location in addition to the effects of age, sex and vascular risk factors
The recent discovery that the prevalence of cysteine mutations in the NOTCH3 gene responsible for CADASIL was more than 100 times higher in the general population than that estimated in patients highlighted that the mutation location in EGFr-like-domains of the NOTCH3 receptor could have a major effect on the phenotype of the disease. The exact impact of such mutations locations on the multiple facets of the disease has not been fully evaluated. We aimed to describe the phenotypic spectrum of a large population of CADASIL patients and to investigate how this mutation location influenced various clinical and imaging features of the disease. Both a supervised and a non-supervised approach were used for analysis. The results confirmed that the mutation location is strongly related to clinical severity and showed that this effect is mainly driven by a different development of the most damaging ischemic tissue lesions at cerebral level. These effects were detected in addition to those of aging, male sex, hypertension and hypercholesterolemia. The exact mechanisms relating the location of mutations along the NOTCH3 receptor, the amount or properties of the resulting NOTCH3 products accumulating in the vessel wall, and their final consequences at cerebral level remain to be determined.
153
Impulse Noise Removal Using Adaptive Radial Basis Function Interpolation
A novel adaptive radial basis function interpolation-based impulse noise removal algorithm is introduced in this manuscript. This approach consists of two stages: noisy pixel detection and correction. In former step, the noise-affected pixels in an image are detected, and in the latter step, the noisy pixels are restored by adaptive radial basis function-based interpolation scheme. The radial basis function interpolation scheme is used to estimate the unknown noisy pixel value from the noise-free known neighboring pixel values. For both noisy pixel detection and correction, a center sliding window is considered at each pixel location. The proposed approach is experimented on some benchmark data sets, and its performance is evaluated using five performance evaluation measures: PSNR, MSSIM, IEF, correlation factor, and NSER on different test images by comparing it against sixteen different state-of-the-art techniques. It is found that the proposed approach gives better results than the sixteen different state-of-the-art techniques.
154
GaitCopy: Disentangling Appearance for Gait Recognition by Signature Copy
This paper addresses the problem of gait-based people identification by copying optical flow-based signatures. The proposed model, coined as GaitCopy, receives as input a stack of gray images and returns the gait signature of the represented subject. The novel property of this network is that it is not trained to only generate discriminative signatures, but to copy signatures generated by a Master network trained on optical flow inputs. Then, GaitCopy is enforced to extract signatures based on motion and not based on appearance, despite having been trained with pixel inputs. We implement two different versions of GaitCopy, one mainly composed of 3D convolutional layers to capture local temporal information; and a second one based on GaitSet which uses 2D convolutional layers under a temporal setup. We evaluate our approach on two public gait datasets: CASIA-B and TUM-GAID. We observe that compact networks, up to x4:2 smaller for TUM-GAID, can be obtained by using our approach, while keeping a competitive recognition accuracy with respect to the state of the art, and without the need of explicit optical flow computation. Even with such network compression, the results obtained in TUM-GAID are comparable to those of the state of the art, with an average accuracy of 97% on the test set. (Code will be publicly available upon acceptance.)
155
Classification of user competency levels using EEG and convolutional neural network in 3D modelling application
Competency classification is one of the main challenging tasks for the development of state-of-the-art next generation computer-aided design (CAD) system. To develop a futuristic system that can accommodate the lack of competency, the system needs to adapt to the competency level of the user. To solve this problem, we have presented a deep convolutional neural network (CNN) model that uses the Electroencephalography (EEG) of the user to classify the level of competency in 3D modeling task. The five competency levels were defined based on the task completion time, final 3D model rating and previous modeling experience. This is the first study that classifies user competency and employs the CNN model for the analysis of EEG signals in the design application. In this work, a 14-layer deep CNN model was implemented to classify competency into five different levels. The proposed technique achieved an accuracy, specificity, and sensitivity of > 88%, > 90% and > 70% respectively with 5-fold cross-validation. The results showed the applicability of a CNN model to classify the user competency and can be used as a first step in developing state-of-the-art adaptive 3D modeling systems. (C) 2020 Elsevier Ltd. All rights reserved.
156
FQI: feature-based reduced-reference image quality assessment method for screen content images
In this study, a reduced-reference image-quality-assessment (IQA) method for screen content images, named as feature-quality-index (FQI) is proposed. The proposed method is based on the fact that the human visual system is more sensitive towards change in features than intensity or structure. Reduced features from the reference and distorted images are first extracted. In order to find the preserved features in the distorted image, a feature matching process with a reduced number of distance calculations is proposed, namely reduced-distance method. To reflect the importance of the matched features and their distance, the inner product between the normalised scale and distance vector is obtained. Extensive comparisons are performed on two available benchmark databases namely SIQAD and QACS, with eight reduced-reference, and nine full-reference state-of-the-art IQA techniques to demonstrate the consistency, accuracy, and robustness of the proposed FQI. The subjective evaluation of mean opinion score shows that FQI outperforms the current state-of-the-art IQA techniques.
157
Ori-Finder 2022: A Comprehensive Web Server for Prediction and Analysis of Bacterial Replication Origins
The replication of DNA is a complex biological process that is essential for life. Bacterial DNA replication is initiated at genomic loci referred to as replication origins (oriCs). Integrating the Z-curve method, DnaA box distribution, and comparative genomic analysis, we developed a web server to predict bacterial oriCs in 2008 called Ori-Finder, which contributes to clarify the characteristics of bacterial oriCs. The oriCs of hundreds of sequenced bacterial genomes have been annotated in the genome reports using Ori-Finder and the predicted results have been deposited in DoriC, a manually curated database of oriCs. This has facilitated large-scale data mining of functional elements in oriCs and strand-biased analysis. Here, we describe Ori-Finder 2022 with updated prediction framework, interactive visualization module, new analysis module, and user-friendly interface. More species-specific indicator genes and functional elements of oriCs are integrated into the updated framework, which has also been redesigned to predict oriCs in draft genomes. The interactive visualization module displays more genomic information related to oriCs and their functional elements. The analysis module includes regulatory protein annotation, repeat sequence discovery, homologous oriC search, and strand-biased analyses. The redesigned interface provides additional customization options for oriC prediction. Ori-Finder 2022 is freely available at http://tubic.tju.edu.cn/Ori-Finder/ and https://tubic.org/Ori-Finder/.
158
State of the art of smart homes
In this paper we present a review of the state of the art of smart homes. We will first look at the research work related to smart homes from various view points; first in the view point of specific techniques such as smart homes that utilize computer vision based techniques, smart homes that utilize audio-based techniques and then smart homes that utilize multimodal techniques. Then we look at it from the view point of specific applications of smart homes such as eldercare and childcare applications, energy efficiency applications and finally in the research directions of multimedia retrieval for ubiquitous environments. We will summarize the smart homes based research into these two categories. In the survey we found out that some well-known smart home applications like video based security applications has seen the maturity in terms of new research directions while some topics like smart homes for energy efficiency and video summarization are gaining momentum. (C) 2012 Elsevier Ltd. All rights reserved.
159
Recent advances in phosphorescent OLEDs for small- and large-area-display sizes
State-of-the-art phosphorescent organic light-emitting diode (PHOLED(TM)) lifetime and efficiency performances for a range of emission colors are reported. Lifetimes in excess of 100,000 hours were demonstrated at display luminance levels for yellow-green and NTSC deep-red emission. In addition, external quantum efficiencies close to the theoretical maximum are demonstrated for long-lived PHOLEDs.
160
Systematic Review of In Vitro Antimicrobial Susceptibility Testing for Bacillus anthracis, 1947-2019
Bacillus anthracis, the causative agent of anthrax, is a high-consequence bacterial pathogen that occurs naturally in many parts of the world and is considered an agent of biowarfare or bioterrorism. Understanding antimicrobial susceptibility profiles of B. anthracis isolates is foundational to treating naturally occurring outbreaks and to public health preparedness in the event of an intentional release. In this systematic review, we searched the peer-reviewed literature for all publications detailing antimicrobial susceptibility testing of B. anthracis. Within the set of discovered articles, we collated a subset of publications detailing susceptibility testing that followed standardized protocols for Food and Drug Administration-approved, commercially available antimicrobials. We analyzed the findings from the discovered articles, including the reported minimal inhibitory concentrations. Across the literature, most B. anthracis isolates were reported as susceptible to current first-line antimicrobials recommended for postexposure prophylaxis and treatment. The data presented for potential alternative antimicrobials will be of use if significant resistance to first-line antimicrobials arises, the strain is bioengineered, or first-line antimicrobials are not tolerated or available.
161
The increased inter-brain neural synchronization in prefrontal cortex between simulated patient and acupuncturist during acupuncture stimulation: Evidence from functional near-infrared spectroscopy hyperscanning
The patient-acupuncturist interaction was a critical influencing factor for acupuncture effects but its mechanism remains unclear. This study aimed to examine the inter-brain mechanism of patient-acupuncturist dyad during acupuncture stimulation in a naturalistic clinical setting. Seventy healthy subjects (simulated "patients") were randomly assigned to two groups and received verum acupuncture group or sham acupuncture by one acupuncturist. Functional near-infrared spectroscopy hyperscanning was used to simultaneously record the neural responses of "patient"-acupuncturist dyad during acupuncture stimulation in each group. The results showed that inter-brain neural synchronization (INS) in the prefrontal cortex (PFC) of "patient"-acupuncturist dyad was significantly increased during verum but not sham acupuncture stimuli, and positively correlated with the needling sensations of "patients." Granger causality analysis demonstrated that there were no significant differences in INS direction between the "patient" and the acupuncturist. This study identified the increase of INS between "patient" and acupuncturist, and suggested that PFC was important to the interaction of "patient"-acupuncturist dyad.
162
Defensive Strongholds and Fortified Castles in Poland-From the Art of Fortifications to Tourist Attractions
The scientific problem undertaken is the importance of castles for the functioning of cultural tourism in the opinion of the inhabitants of Central Europe. What is the use of medieval monuments for the art of fortifications today? The main method of research is a diagnostic survey carried out with the use of a survey on a group of N, important according to the statistics in the number of n = 614 respondents. Statistical analyzes were performed using Statistica version 13.3. On the basis of the presented research results, it can be concluded that the interests of the respondents are very broad and varied. Taking into account the relatively large group of respondents, the research results can be considered reliable. An important goal was supplementing the knowledge, meeting the needs of learning about history, and getting acquainted with the prevailing historical tradition in castles of Europe. The questions presented here accent the interests of castles for the functioning of cultural tourism in the opinion of the inhabitants of Central Europe.
163
Rock art provides new evidence on the biogeography of kudu (Tragelaphus imberbis), wild dromedary, aurochs (Bos primigenius) and African wild ass (Equus africanus) in the early and middle Holocene of north-western Arabia
Aim: Our knowledge of the prehistoric distribution of animal species is so far largely dependent on the location of excavated archaeological and palaeontological sites. In the absence of excavated faunal remains, many species that were present in the Levant and North Africa have been assumed to have been absent on the Arabian Peninsula. Here, we explore representations of four species that were identifiable in the rock art, but had not previously been reported in north-western Arabia. Location: Jubbah and Shuwaymis UNESCO world heritage rock art sites in Ha'il province, north-western Saudi Arabia. Methods: In total, the rock art panels surveyed and recorded in Jubbah and Shuwaymis contain 6,618 individual animal depictions. Species were identified based on diagnostic features of the anatomy. The resulting dataset was then compared to the faunal spectrum reported in the (archaeo) zoological literature. Results: The rock art dataset provides evidence that the distributions of lesser kudu (Tragelaphus imberbis), wild camel and African wild ass (Equus africanus) extended into the north-west of Arabia and that the engravers may have had knowledge of aurochs (Bos primigenius). Main conclusions: The presence of previously undocumented mammal species in Arabia provides new information regarding their distribution, as well as the types of habitat and vegetation that were available in prehistoric landscapes. Moreover, the presence of kudu on the Arabian Peninsula indicates that the identification of palaeo-distributions based exclusively on faunal remains may miss key species in the Afro-Eurasian faunal exchange.
164
Aeronautical Networks for In-Flight Connectivity: A Tutorial of the State-of-the-Art and Survey of Research Challenges
The aeronautical networks attract the attention of both industry and academia since Internet access during flights turns to the crucial demand from luxury with the evolving technology. This In-Flight Connectivity (IFC) necessity is currently dominated by the satellite connectivity and Air-to-Ground (A2G) network solutions. However, the high installation/equipment cost and latency of the satellite connectivity reduce its efficiency. The A2G networks are utilized through the 4G/5G ground stations deployed on terrestrial areas to solve these satellites' problems. This terrestrial deployment reduces the coverage area of A2G networks, especially for remote flights over the ocean. The Aeronautical Ad-hoc Networks (AANETs) are designed to provide IFC while solving the primary defects of dominating solutions. The AANET is an entirely novel solution under the vehicular networks since it consists of aircraft with ultra-dynamic and unstructured characteristics. These characteristics separate it from the less dynamic Flying Ad-Hoc Networks (FANETs). Therefore, the environmental and mobility effects cause specific challenges for AANETs. This article presents a holistic review of these open AANET challenges by investigating them in data link, network, and transport layers. Before giving the details of these challenges, this article explores the state-of-the-art literature about satellite and A2G networks for IFC. We then give our specific interest to the AANET by investigating its particular characteristics and open research challenges. The main starting point of this study is that there is a lack of compact research on this exciting topic, although IFC is an inevitable need for the aeronautical industry. Also, the AANET could be underlined by giving all state-of-the-art about the dominating IFC solutions. Therefore, this is the first work exploring the state-of-the-art for all the existing aeronautical networking technologies under a single comprehensive survey by deeply analyzing specific characteristics and open research challenges of AANETs. Additionally, the AANET is a novel topic and should be separately investigated from the FANETs as given in current literature.
165
HyperDense-Net: A Hyper-Densely Connected CNN for Multi-Modal Image Segmentation
Recently, dense connections have attracted substantial attention in computer vision because they facilitate gradient flow and implicit deep supervision during training. Particularly, DenseNet that connects each layer to every other layer in a feed-forward fashion and has shown impressive performances in natural image classification tasks. We propose HyperDenseNet, a 3-D fully convolutional neural network that extends the definition of dense connectivity to multi-modal segmentation problems. Each imaging modality has a path, and dense connections occur not only between the pairs of layers within the same path but also between those across different paths. This contrasts with the existing multi-modal CNN approaches, in which modeling several modalities relies entirely on a single joint layer (or level of abstraction) for fusion, typically either at the input or at the output of the network. Therefore, the proposed network has total freedom to learn more complex combinations between the modalities, within and in-between all the levels of abstraction, which increases significantly the learning representation. We report extensive evaluations over two different and highly competitive multi-modal brain tissue segmentation challenges, iSEG 2017 and MRBrainS 2013, with the former focusing on six month infant data and the latter on adult images. HyperDenseNet yielded significant improvements over many state-of-the-art segmentation networks, ranking at the top on both benchmarks. We further provide a comprehensive experimental analysis of features re-use, which confirms the importance of hyperdense connections in multi-modal representation learning. Our code is publicly available.
166
Minimum Mean-Square Error Estimation of Mel-Frequency Cepstral Features-A Theoretically Consistent Approach
In this work, we consider the problem of feature enhancement for noise-robust automatic speech recognition (ASR). We propose a method for minimum mean-square error (MMSE) estimation of mel-frequency cepstral features, which is based on a minimum number of well-established, theoretically consistent statistical assumptions. More specifically, the method belongs to the class of methods relying on the statistical framework proposed in Ephraim and Malah's original work ("Speech enhancement using a minimum mean-square error short-time spectral amplitude estimator," IEEE Trans. Acoust., Speech, Signal Process., vol. ASSP-32, no. 6, 1984). The method is general in that it allows MMSE estimation of mel-frequency cepstral coefficients (MFCC's), cepstral-mean subtracted (CMS-) MFCC's, autoregressive-moving-average (ARMA)-filtered CMS-MFCC's, velocity, and acceleration coefficients. In addition, the method is easily modified to take into account other compressive non-linearities than the logarithm traditionally used for MFCC computation. In terms of MFCC estimation performance, as measured by MFCC mean-square error, the proposed method shows performance which is identical to or better than other state-of-the-art methods. In terms of ASR performance, no statistical difference could be found between the proposed method and the state-of-the-art methods. We conclude that existing state-of-the-art MFCC feature enhancement algorithms within this class of algorithms, while theoretically suboptimal or based on theoretically inconsistent assumptions, perform close to optimally in the MMSE sense.
167
Optimisation of Deep Learning Small-Object Detectors with Novel Explainable Verification
In this paper, we present a novel methodology based on machine learning for identifying the most appropriate from a set of available state-of-the-art object detectors for a given application. Our particular interest is to develop a road map for identifying verifiably optimal selections, especially for challenging applications such as detecting small objects in a mixed-size object dataset. State-of-the-art object detection systems often find the localisation of small-size objects challenging since most are usually trained on large-size objects. These contain abundant information as they occupy a large number of pixels relative to the total image size. This fact is normally exploited by the model during training and inference processes. To dissect and understand this process, our approach systematically examines detectors' performances using two very distinct deep convolutional networks. The first is the single-stage YOLO V3 and the second is the double-stage Faster R-CNN. Specifically, our proposed method explores and visually illustrates the impact of feature extraction layers, number of anchor boxes, data augmentation, etc., utilising ideas from the field of explainable Artificial Intelligence (XAI). Our results, for example, show that multi-head YOLO V3 detectors trained using augmented data produce better performance even with a fewer number of anchor boxes. Moreover, robustness regarding the detector's ability to explain how a specific decision was reached is investigated using different explanation techniques. Finally, two new visualisation techniques are proposed, WS-Grad and Concat-Grad, for identifying explanation cues of different detectors. These are applied to specific object detection tasks to illustrate their reliability and transparency with respect to the decision process. It is shown that the proposed techniques can result in high resolution and comprehensive heatmaps of the image areas, significantly affecting detector decisions as compared to the state-of-the-art techniques tested.
168
Multipath Scheduling for 5G Networks: Evaluation and Outlook
The fifth generation (5G) of cellular networks aims at providing very high data rates, ultra-reliable low latency, and massive connection density. As one of the fundamental design trends toward these objectives, 5G exploits multi-connectivity (i.e., the concurrent use of multiple access networks), where multipath transport protocols have emerged as key technology enablers. The scheduler of a multipath transport protocol determines how to distribute the data packets onto different paths and has a critical impact on the protocol performance. Within this context, we present in this article the first empirical evaluation of state-of-the-art multipath schedulers based on real 5G data, for both static and mobile scenarios. Furthermore, we introduce M-Peekaboo, which builds on a state-of-the-art learning-based multipath scheduler and extends its usage to 5G networks. Our results illustrate the benefits of employing a learning-based multipath scheduler for 5G networks and motivate further studies of advanced learning schemes that can adapt more quickly to the path conditions, and take into account the emerging features and requirements of 5G and beyond networks.
169
From genotypes to phenotypes: expression levels of genes encompassing adaptive SNPs in black spruce
Measuring transcript levels for adaptive genes revealed polymorphisms having cis -effect upon gene expression levels related to phenotype variation in a black spruce natural population. Trees growing in temperate and boreal regions must acclimate to changes in climatic factors such as low winter temperatures to survive to seasonal variations. Common garden studies have shown that genetic variation in quantitative traits helps species to survive and adapt to environmental changes and local conditions. Twenty-four genes carrying SNPs were previously associated with genetic adaptation in black spruce (Picea mariana [Mill.] BSP). The objectives of this study were to investigate the potential role of these genes in regulation of winter acclimation and adaptation by studying their patterns of expression as a function of the physiological stage during the annual growth cycle, tissue type, and their SNP genotypic class. Considerable variability in gene expression was observed between different vegetative tissues or organs, and between physiological stages. The genes were expressed predominantly in tissues that could be linked more directly to winter acclimation and adaptation. The expression levels of several of the genes were significantly related to variation in tree height growth or budset timing and expression level variation related to SNP genotypic classes was observed in four of the genes. An interaction between genotypic classes and physiological stages was also observed for some genes, indicating genotypes with different reaction norms in terms of gene expression.
170
Health Outcomes of Kansas City's Vulnerable Patients Following Shutdown: An Assessment of Blood Pressure Among Sojourner Health Clinic Patients
Background In downtown Kansas City, patients who face homelessness or unstable housing situations may have been negatively affected by the shutdown of Sojourner Health Clinic (SOJO), a free student-run clinic that provides primary care predominantly to these patients. Research shows that blood pressure (BP) increases within weeks or months of interruption of antihypertensive therapy, especially in patients with advanced age and polypharmacy. Therefore, this study will examine how patients' blood pressure changed after the closure of Sojourner Health Clinic. Methods The study population consists of Sojourner Health Clinic patients who were seen both before March 2020 (shutdown) and during/after July 2020 (clinic reopening). Participants are selected at random. No additional data is collected outside of routine treatment for this institutional review board (IRB)-exempt project. A study coordinator reviews charts via Sojourner electronic medical record (EMR) and collects the latest BP available before March 2020 and the first BP available during/after July 2020. No identifying information is collected. The mean systolic pressures, mean diastolic pressures, and mean arterial pressures (MAP) are compared via paired t-test for statistical significance. Results There was a statistically significant decrease in patients' MAP and diastolic BP after the closure of the clinic. However, there was not a statistically significant change found in patients' systolic BP. The clinical significance of these results is limited by the minimal magnitude of change. Conclusions These findings run counter to our expectations since we believed that the closure of Sojourner Health Clinic would correlate with worsened markers of health, such as blood pressure control. It may be possible that the sampled patients turned to other sources for health maintenance and antihypertensive therapy during clinic closure. Future studies could explore these possibilities.
171
Scene wireframes sketching for Unmanned Aerial Vehicles
This paper introduces novel insights to improve the state-of-the-art line-based unsupervised observation and abstraction models of man-made environments. The increasing use of autonomous UAVs inside buildings and around human-made structures demands new accurate and comprehensive representation of their operation environments. Most of the 3D scene abstraction methods use invariant feature point matching, nevertheless some sparse 3D point clouds do not concisely represent the structure of the environment. The presented approach is based on observation and representation models using the straight line segments. The goal of the work is a complete method based on the matching of lines, that provides a complementary approach to state-of-the-art methods when facing 3D scene representation of poor texture environments for future autonomous UAV. Oppositely to other recently published methods obtaining 3D line abstractions, the proposed method features 3D segment abstraction in the absence of a previously generated point based reconstruction. Another advantage is the ability to group the resulting 3D lines according to different planes, for exploiting coplanar line intersections. These intersections are used like feature points in the reconstruction process. It has been proved that this method exclusively based on lines can obtain spatial information in the adverse situations when a SIFT-like SfM pipeline fails to generate a dense point cloud. (C) 2018 Elsevier Ltd. All rights reserved.
172
Microbial production of lactic acid: the latest development
Lactic acid is an important platform chemical for producing polylactic acid (PLA) and other value-added products. It is naturally produced by a wide spectrum of microbes including bacteria, yeast and filamentous fungi. In general, bacteria ferment C5 and C6 sugars to lactic acid by either homo- or hetero-fermentative mode. Xylose isomerase, phosphoketolase, transaldolase, l- and d-lactate dehydrogenases are the key enzymes that affect the ways of lactic acid production. Metabolic engineering of microbial strains are usually needed to produce lactic acid from unconventional carbon sources. Production of d-LA has attracted much attention due to the demand for producing thermostable PLA, but large scale production of d-LA has not yet been commercialized. Thermophilic Bacillus coagulans strains are able to produce l-lactic acid from lignocellulose sugars homo-fermentatively under non-sterilized conditions, but the lack of genetic tools for metabolically engineering them severely affects their development for industrial applications. Pre-treatment of agriculture biomass to obtain fermentable sugars is a pre-requisite for utilization of the huge amounts of agricultural biomass to produce lactic acid. The major challenge is to obtain quality sugars of high concentrations in a cost effective-way. To avoid or minimize the use of neutralizing agents during fermentation, genetically engineering the strains to make them resist acidic environment and produce lactic acid at low pH would be very helpful for reducing the production cost of lactic acid.
173
Data-Driven Online Speed Optimization in Autonomous Sailboats
This paper addresses the issue of data-driven online velocity optimization of an autonomous sailboat. Autonomous sailboats represent an ideal for long range and duration reconnaissance missions. Sailboat control is a challenging control task; sailboats are characterized by a number of control variables, all of which affect the ship trajectory and state in a highly nonlinear fashion. In this paper, a path-following automatic sailboat controller is presented. The control system has two main components: a heading control, acting on the rudder, and a velocity optimizer, acting on the sails. The optimizer is based on a modified extremum seeking approach. This paper also derives a first-principle-based 4 DoF sailboat model that is experimentally validated and used to guide the design and tuning of the control system. In fact, the control system is first tuned and validated in simulation. The simulation environment enables the comparison of the proposed model against a theoretical benchmark and a state-of-the-art controller. The analysis reveals that the proposed control system achieves near-optimal performance and considerably outperforms the state-of-the-art solution. Finally, the controller is tested and validated on an instrumented scale model.
174
The dynamics of mercury around an artisanal and small-scale gold mining area, Camarines Norte, Philippines
To elucidate the dynamics of mercury emitted and released by artisanal and small-scale gold mining (ASGM) activity and to estimate its impact on the ecosystems of the bay, the distribution of mercury in the atmosphere, soil, water, and sediment around Mambulao Bay, Camarines Norte, Philippines, was investigated. The ASGM operations use mercury to extract gold from ore and are located on the east shore side of the bay. Samplings were conducted in August 2017 and September 2018. The samples were used for determination of total mercury (T-Hg) and organic mercury (org-Hg) concentrations, total organic carbon (TOC) content, and chemical composition. The atmospheric mercury concentration on the east shore side, 6.1-25.8 ng m-3, was significantly higher than the value of 1.4-9.9 ng m-3 observed on the west shore side. The average concentrations of T-Hg in the forest soils of the west shore side and those of the east shore side were 0.081 ± 0.028 mg kg-1 and 0.496 ± 0.439 mg kg-1, respectively. In the vertical distribution of T-Hg in the soil of the east shore side, a higher concentration was observed near the surface. For the vertical variations in T-Hg in the marine sediment, higher values were observed near the estuary, and the vertical variations in core samples showed an increase in mercury concentration toward the surface. The highest concentration of T-Hg in sediment, 9.5 mg kg-1, which was 2 orders of magnitude higher than the background levels of this area, was found near the river mouth. The T-Hg, org-Hg, and TOC levels showed a positive correlation, suggesting that the rivers are the main sources of T-Hg and org-Hg in the bay. Although the fish sample containing a mercury content higher than the regulatory level for fish and shellfish of 0.4 mg kg-1 in Japan was only one of 42 samples, the percentage of org-Hg in fish samples was 91 ± 18%. Mercury released into the surroundings by the ASGM activities can be converted into methylmercury and affect the bay's ecosystem.
175
Image Registration Based on Low Rank Matrix: Rank-Regularized SSD
Similarity measure is a main core of image registration algorithms. Spatially varying intensity distortion is an important challenge, which affects the performance of similarity measures. Correlation among the pixels is the main characteristic of this distortion. Similarity measures such as sum-of-squared-differences(SSD) and mutual information ignore this correlation; hence, perfect registration cannot be achieved in the presence of this distortion. In this paper, we model this correlation with the aid of the low rank matrix theory. Based on this model, we compensate this distortion analytically and introduce rank-regularized SSD (RRSSD). This new similarity measure is a modified SSD based on singular values of difference image in mono-modal imaging. In fact, image registration and distortion correction are performed simultaneously in the proposed model. Based on our experiments, the RRSSD similarity measure achieves clinically acceptable registration results, and outperforms other state-of-the-art similarity measures, such as the well-known method of residual complexity.
176
Two-Stream Graph Convolutional Network for Intra-Oral Scanner Image Segmentation
Precise segmentation of teeth from intra-oral scanner images is an essential task in computer-aided orthodontic surgical planning. The state-of-the-art deep learning-based methods often simply concatenate the raw geometric attributes (i.e., coordinates and normal vectors) of mesh cells to train a single-stream network for automatic intra-oral scanner image segmentation. However, since different raw attributes reveal completely different geometric information, the naive concatenation of different raw attributes at the (low-level) input stage may bring unnecessary confusion in describing and differentiating between mesh cells, thus hampering the learning of high-level geometric representations for the segmentation task. To address this issue, we design a two-stream graph convolutional network (i.e., TSGCN), which can effectively handle inter-view confusion between different raw attributes to more effectively fuse their complementary information and learn discriminative multi-view geometric representations. Specifically, our TSGCN adopts two input-specific graph-learning streams to extract complementary high-level geometric representations from coordinates and normal vectors, respectively. Then, these single-view representations are further fused by a self-attention module to adaptively balance the contributions of different views in learning more discriminative multi-view representations for accurate and fully automatic tooth segmentation. We have evaluated our TSGCN on a real-patient dataset of dental (mesh) models acquired by 3D intraoral scanners. Experimental results show that our TSGCN significantly outperforms state-of-the-art methods in 3D tooth (surface) segmentation.
177
Recursive estimation for sparse Gaussian process regression
Gaussian Processes (GPs) are powerful kernelized methods for non-parametric regression used in many applications. However, their use is limited to a few thousand of training samples due to their cubic time complexity. In order to scale GPs to larger datasets, several sparse approximations based on so-called inducing points have been proposed in the literature. In this work we investigate the connection between a general class of sparse inducing point GP regression methods and Bayesian recursive estimation which enables Kalman Filter like updating for online learning. The majority of previous work has focused on the batch setting, in particular for learning the model parameters and the position of the inducing points, here instead we focus on training with mini-batches. By exploiting the Kalman filter formulation, we propose a novel approach that estimates such parameters by recursively propagating the analytical gradients of the posterior over mini-batches of the data. Compared to state of the art methods, our method keeps analytic updates for the mean and covariance of the posterior, thus reducing drastically the size of the optimization problem. We show that our method achieves faster convergence and superior performance compared to state of the art sequential Gaussian Process regression on synthetic GP as well as real-world data with up to a million of data samples. (C) 2020 Elsevier Ltd. All rights reserved.
178
Continuous face authentication scheme for mobile devices with tracking and liveness detection
We present a novel scheme for continuous face authentication using mobile device cameras that addresses the issue of spoof attacks and attack windows in state-of-the-art approaches. Our scheme authenticates a user based on extracted facial features. However, unlike other schemes that periodically re-authenticate a user, our scheme tracks the authenticated face and only attempts re-authentication when the authenticated face is lost. This allows our scheme to eliminate attack windows that exist in schemes authenticating periodically and immediately recognise impostor usage. We also introduce a robust liveness detection component to our scheme that can detect printed faces and face videos. We describe how the addition of liveness detection enhances the robustness of our scheme against spoof attacks, improving on state-of-the-art approaches that lack this capability. Furthermore, we create the first dataset of facial videos collected from mobile devices during different real-world activities (walking, sitting and standing) such that our results reflect realistic scenarios. Our dataset therefore allows us to give new insight into the impact of user activity on facial recognition. Our dataset also includes spoofed facial videos for liveness testing. We use our dataset alongside two benchmark datasets for our experiments. We show and discuss how our scheme improves on existing continuous face authentication approaches and efficiently enhances device security.
179
Circular Clustering in Fuzzy Approximation Spaces for Color Normalization of Histological Images
One of the foremost and challenging tasks in hematoxylin and eosin stained histological image analysis is to reduce color variation present among images, which may significantly affect the performance of computer-aided histological image analysis. In this regard, the paper introduces a new rough-fuzzy circular clustering algorithm for stain color normalization. It judiciously integrates the merits of both fuzzy and rough sets. While the theory of rough sets deals with uncertainty, vagueness, and incompleteness in stain class definition, fuzzy set handles the overlapping nature of histochemical stains. The proposed circular clustering algorithm works on a weighted hue histogram, which considers both saturation and local neighborhood information of the given image. A new dissimilarity measure is introduced to deal with the circular nature of hue values. Some new quantitative measures are also proposed to evaluate the color constancy after normalization. The performance of the proposed method, along with a comparison with other state-of-the-art methods, is demonstrated on several publicly available standard data sets consisting of hematoxylin and eosin stained histological images.
180
A Study on the Factors Affecting Academic Achievement in the Non-Face-to-Face Class Environment Due to COVID-19: Focusing on Computer Liberal Arts Education Class
As a result of the COVID-19 pandemic, many universities have shifted to non-face-to-face classes resulting in numerous changes in the educational system. Since programming education includes a greater proportion of practice than theory-oriented courses, non-face-to-face classes have several constraints. As a result, to properly execute software education and enhance educational performance for non-major students, it is required to conduct research. Students' psychological moods and activities collected in online classrooms were used to investigate factors impacting academic success as measured by scores and grades. Multiple regression analysis and logistic regression analysis were conducted by using data mining analytical approach. Attendance, effort expectancy, hedonic motivation, confidence, frequency of communication in mobile chat rooms, and Python programming intention factors were retrieved as an outcome of the performance. The relevance of the factors was confirmed, and it was revealed that hedonic motivation was crucial for students in Class A, while attendance had a significant impact on academic progress for students in the other grades. The goal of this research is to assist university organizations in making decisions by enhancing computer liberal arts education and offering implications for future non-face-to-face teaching environments such as during the COVID-19 pandemic.
181
ST-TrafficNet: A Spatial-Temporal Deep Learning Network for Traffic Forecasting
This paper presents a spatial-temporal deep learning network, termed ST-TrafficNet, for traffic flow forecasting. Recent deep learning methods highly relate accurate predetermined graph structure for the complex spatial dependencies of traffic flow, and ineffectively harvest high dimensional temporal features of the traffic flow. In this paper, a novel multi-diffusion convolution block constructed by an attentive diffusion convolution and bidirectional diffusion convolution is proposed, which is capable to extract precise potential spatial dependencies. Moreover, a stacked Long Short-Term Memory (LSTM) block is adopted to capture high-dimensional temporal features. By integrating the two blocks, the ST-TrafficNet can learn the spatial-temporal dependencies of intricate traffic data accurately. The performance of the ST-TrafficNet has been evaluated on two real-world benchmark datasets by comparing it with three commonly-used methods and seven state-of-the-art ones. The Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE) of the proposed method outperform not only the commonly-used methods, but also the state-of-the-art ones in 15 min, 30 min, and 60 min time-steps.
182
Development of a comprehensive two-dimensional liquid chromatographic mass spectrometric method for the non-targeted identification of poly- and perfluoroalkyl substances in aqueous film-forming foams
In this research, we developed an online comprehensive two-dimensional liquid chromatographic (LC × LC) method hyphenated with high-resolution mass spectrometry (HRMS) for the non-targeted identification of poly- and perfluorinated compounds (PFASs) in fire-fighting aqueous-film forming foams (AFFFs). The method exploited the combination of mixed-mode weak anion exchange-reversed phase with a octadecyl stationary phase, separating PFASs according to ionic classes and chain length. To develop and optimize the LC × LC method we used a reference training set of twenty-four anionic PFASs, representing the main classes of compounds occurring in AFFFs and covering a wide range of physicochemical properties. In particular, we investigated different modulation approaches to reduce injection band broadening and breakthrough in the second dimension separation. Active solvent and stationary phase assisted modulations were compared, with the best results obtained with the last approach. In the optimal conditions, the predicted peak capacity corrected for undersampling was higher than three-hundred in a separation space of about 60 min. Subsequently, the developed method was applied to the non-targeted analysis of two AFFF samples for the identification of homologous series of PFASs, in which it was possible to identify up to thirty-nine potential compounds of interest utilizing Kendrick mass defect analysis. Even within the samples, the features considered potential PFAS by mass defect analysis elute in the chromatographic regions discriminating for the ionic group and/or the chain length, thus confirming the applicability of the method presented for the analysis of AFFF mixtures and, to a further extent, of environmental matrices affected by the AFFF.
183
Early ontogeny of the commercially valuable fish red-bellied pacu Piaractus brachypomus (Characiformes, Serrasalmidae) from the Amazon, Brazil
The initial development of the red-bellied pacu Piaractus brachypomus is described using morphological, meristic and morphometric characteristics. A total of 127 individuals were analysed (47 in the yolk-sac, 35 in pre-flexion, five in flexion, 20 in post-flexion and 20 in juvenile) with standard length varying between 2.92 and 48.61 mm. The larvae are born poorly developed and have a discoidal yolk at ~6.33 mm standard length. During early ontogeny, the mouth passes from terminal to subterminal and the anal opening reaches the vertical line over the midline region of the body. There are changes in body shape from long and moderate to deep, head length from small to large, and eye diameter from moderate to large. Dendritic chromatophores were present in the ventral, dorsal and upper part of the swim bladder in the early larval stages. Rounded spots are evident all over the body in juveniles. The total number of myomeres ranges from 39 to 41 (20-23 pre-anal, 17-20 post-anal). Through the morphometric relationships, it was evidenced that the greatest changes during the initial ontogeny of P. brachypomus occur in the transition from the post-flexion stage to the juvenile period, indicating changes in behaviour, foraging and physiology.
184
Nanoemulsion-templated polylelectrolyte multifunctional nanocapsules for DNA entrapment and bioimaging
The emerging field of bionanotechnology aims at advancing colloidal and biomedical research via introduction of multifunctional nanoparticle-based containers intended for both gene therapy and bioimaging. In the present contribution we entrapped the model genetic material (herring testes DNA) in the newly-designed non-viral vectors, i.e., multifunctional nanocapsules obtained by layer-by-layer (LbL) adsorption of DNA and oppositely charged polysaccharide-based chitosan (CHIT) on the nanoemulsion core, loaded by IR-780 indocyanine (used as the fluorescent marker) and stabilized by gemini-type ammonium salts: N,N,N',N'-tetramethyl-N,N'-di(dodecyl)-ethylenediammonium bromide, d(DDA)PBr and N,N,N',N'-tetramethyl-N,N'-di(dodecyl)-butylenediammonium d(DDA)BBr. Ternary phase diagrams of the surfactant-oil-water systems were determined by titration method. Then, the stability of the nanoemulsions obtained with IR-780 solubilized in the oleic acid (OA) or isopropyl myristate (IPM) phase was evaluated by backscattering (BS) profiles and ζ-potential measurements. In the next step, CHIT and DNA layers were subsequently deposited on the kinetically stable nanoemulsion cores. The IR-780-loaded nanocarriers covered by (DNA/CHIT)4 bilayers shown the high ζ-potential value (about +43mV provided by Doppler electrophoresis), the size <120nm and the spherical shape as analyzed by dynamic light scattering (DLS), atomic force microscopy (AFM) and scanning electron microscopy (SEM). Finally, the long-lasting nanosystems were subjected to in vitro biological studies on human cancer cell lines - doxorubicin-sensitive breast (MCF-7/WT), epithelial lung adenocarcinoma (A549) and skin melanoma (MEWO). Biological response of the cell culture was expressed as cytotoxic activity evaluated by MTT-based proliferation assay as well as bioimaging of intracellular localization of IR-780 molecules loaded in the multilayer DNA-deposited nanocontainers - provided by confocal laser scanning microscopy (CLSM) and total internal reflection fluorescence microscopy (TIRFM). Our results demonstrate that the fabricated oil-core CHIT-coated nanocapsules stabilized by both d(DDA)PBr and d(DDA)BBr surfactants are promising as multifunctional nanocarriers for DNA delivery and cancer diagnostics.
185
NP-hardness of broadcast scheduling and inapproximability of single-source unsplittable min-cost flow
We consider the version of broadcast scheduling where a server can transmit W messages of a given set at each time-step. answering previously made requests for these messages. The goal is to minimize the average response time (ART) if the amount of requests is known in advance for each time-step and message. We prove that this problem is NP-hard, thus answering an open question stated by Kalyanasundaram, Pruhs and Velauthapillai (Proceedings of ESA 2000, LNCS 1879, 2000, pp. 290-301). Furthermore, we present an approximation algorithm that is allowed to send several messages at once. Using six channels for transmissions. the algorithm achieves an ART that is at least Lis good as the optimal solution using one channel. From the NP-hardness of broadcast scheduling we derive a new inapproximability result of (2 - epsilon, 1) for the (congestion. cost) bicriteria version of the single source unsplittable min-cost flow problem, for arbitrary epsilon > 0. The result holds even in the often considered case where the maximum demand is less than or equal to the minimum edge capacity (d(max) less than or equal to u(min)), a case for which an algorithm with ratio (3, 1) was presented by Skutella.
186
Survival pattern of colorectal cancer in Sub-Saharan Africa: A systematic review and meta-analysis
Cancer incidence is relatively low in sub-Saharan Africa (SSA), however, prognosis is expected to be poor in comparison with high-income countries. Comprehensive evidence is limited on the survival pattern of colorectal cancer patients in the region. We conducted a systematic review and meta-analysis to investigate the pattern of colorectal cancer survival in the region and to identify variation across countries and over time. We searched international databases MEDLINE, Scopus, Embase, Web of Science, ProQuest, CINAHL, and Google Scholar to retrieve studies that estimated survival from colorectal cancer in SSA countries from inception to December 31, 2021 without language restriction. Due to between-study heterogeneity, we performed a random-effects meta-analysis to pool survival rates. To identify study-level sources of variation, we performed subgroup analysis and meta-regression. Results are reported in line with the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) 2020 guideline and the protocol was registered in PROSPERO database (CRD42021246935). 23 studies involving 10,031 patients were included in the review, of which, 20 were included in the meta-analysis. The meta-analysis results showed that the pooled 1-, 2-, 3-, 4-, and 5-year survival rates in SSA were 0.74 (95% CI, 0.66-0.81), 0.50 (95% CI, 0.41-0.58), 0.36 (95% CI, 0.27-0.47), 0.31 (95% CI, 0.22-0.42), and 0.28 (95% CI, 0.19-0.38) respectively. Subgroup analyses indicated that the survival rate varied according to year of study, in which those conducted in recent decades showed relatively better survival. The 5-year survival was higher in middle-income SSA countries (0.31; 95%CI: 0.17-0.49) than low-income countries (0.20; 95%CI: 0.11-0.35), however, the difference was not statistically significant. In conclusion, survival from colorectal cancer is low in sub-Saharan Africa compared to other regions. Thus, intervention strategies to improve screening, early diagnosis and treatment of colorectal cancer should be developed and implemented to improve survival in the region.
187
3D Neuron Microscopy Image Segmentation via the Ray-Shooting Model and a DC-BLSTM Network
The morphology reconstruction (tracing) of neurons in 3D microscopy images is important to neuroscience research. However, this task remains very challenging because of the low signal-to-noise ratio (SNR) and the discontinued segments of neurite patterns in the images. In this paper, we present a neuronal structure segmentation method based on the ray-shooting model and the Long Short-Term Memory (LSTM)-based network to enhance the weak-signal neuronal structures and remove background noise in 3D neuron microscopy images. Specifically, the ray-shooting model is used to extract the intensity distribution features within a local region of the image. And we design a neural network based on the dual channel bidirectional LSTM (DC-BLSTM) to detect the foreground voxels according to the voxel-intensity features and boundary-response features extracted by multiple ray-shooting models that are generated in the whole image. This way, we transform the 3D image segmentation task into multiple 1D ray/sequence segmentation tasks, which makes it much easier to label the training samples than many existing Convolutional Neural Network (CNN) based 3D neuron image segmentation methods. In the experiments, we evaluate the performance of our method on the challenging 3D neuron images from two datasets, the BigNeuron dataset and the Whole Mouse Brain Sub-image (WMBS) dataset. Compared with the neuron tracing results on the segmented images produced by other state-of-the-art neuron segmentation methods, our method improves the distance scores by about 32% and 27% in the BigNeuron dataset, and about 38% and 27% in the WMBS dataset.
188
MsTGANet: Automatic Drusen Segmentation From Retinal OCT Images
Drusen is considered as the landmark for diagnosis of AMD and important risk factor for the development of AMD. Therefore, accurate segmentation of drusen in retinal OCT images is crucial for early diagnosis of AMD. However, drusen segmentation in retinal OCT images is still very challenging due to the large variations in size and shape of drusen, blurred boundaries, and speckle noise interference. Moreover, the lack of OCT dataset with pixel-level annotation is also a vital factor hindering the improvement of drusen segmentation accuracy. To solve these problems, a novel multi-scale transformer global attention network (MsTGANet) is proposed for drusen segmentation in retinal OCT images. In MsTGANet, which is based on U-Shape architecture, a novel multi-scale transformer non-local (MsTNL) module is designed and inserted into the top of encoder path, aiming at capturing multi-scale non-local features with long-range dependencies from different layers of encoder. Meanwhile, a novel multi-semantic global channel and spatial joint attention module (MsGCS) between encoder and decoder is proposed to guide the model to fuse different semantic features, thereby improving the model's ability to learn multi-semantic global contextual information. Furthermore, to alleviate the shortage of labeled data, we propose a novel semi-supervised version of MsTGANet (Semi-MsTGANet) based on pseudo-labeled data augmentation strategy, which can leverage a large amount of unlabeled data to further improve the segmentation performance. Finally, comprehensive experiments are conducted to evaluate the performance of the proposed MsTGANet and Semi-MsTGANet. The experimental results show that our proposed methods achieve better segmentation accuracy than other state-of-the-art CNN-based methods.
189
Comparative Appraisal of Leaf Proteomic and Mass Spectrometry Analyses During Fusarium Wilt Infection in Resistance and Susceptible Genotypes of Castor (Ricinus communis L.)
The resistant and susceptible genotypes of castor were utilized for leaf proteomic study during Fusarium wilt infection. The histopathological study was observed under SEM and it confirmed that the infection of Fusarium oxysporum f. sp. ricini was higher in the root of susceptible JI-35, while incompatible interaction is observed in resistant SKI-215 genotype. The acidic and neutral proteins were maximally up-expressed with 2 to 171 kDa in treated resistant and 2 to 150 kDa in treated susceptible interactions. In resistant genotype, the leaf proteins were recognized with 3.0- and 5.8-fold higher at infection stage and post-infection stage, respectively, as compared to susceptible genotype. The highly up expressions of leaf acidic (4.76 pI) and basic (8.77 pI) proteins were found with 224.94- and 61.68-fold change, respectively during the post-infection stage in treated resistance compared to its control. The protein spots at 4.76 pI and 8.77 pI were characterized with nanoLC-MS Triple TOF and were recognized as signalling molecules small GTP binding protein (23 kDa) and actin (8 kDa), respectively, on the basis of mass spectrometry and peptide sequences. However, basic and neutral proteins were up regulated as 30.11- and 20.30-fold changes in treated susceptible compared to its control. These proteins were identified as HSP90 (10 kDa) and LEA (27 kDa) proteins. The 148 kDa protein is recognized as histidine kinase in incompatible resistant interaction compared to compatible susceptible (serine threonine protein kinase, 65 kDa) as common acidic protein at 3.80 pI during infection stage. Some acidic proteins were maximally up-regulated in the leaf of resistant castor genotype and played a significant role in defense response.
190
The relationship between substance use and physical activity among people living with HIV, chronic pain, and symptoms of depression: a cross-sectional analysis
ABSTRACTChronic pain, depression, and substance use are common among people living with HIV (PLWH). Physical activity can improve pain and mental health. Some substances such as cannabis may alleviate pain, which may allow PLWH to participate in more physical activity. However, risks of substance use include poorer mental health and HIV clinical outcomes. This cross-sectional analysis examined the relationships of self-reported substance use (alcohol, cannabis, and nicotine use), gender, and age with self-reports of walking, moderate physical activity, and vigorous physical activity, converted to Metabolic Equivalent of Task Units (METs), among 187 adults living with HIV, chronic pain, and depressive symptoms in the United States. Women reported less walking, vigorous activity, and total physical activity compared to men. Individuals who used cannabis reported more vigorous physical activity relative to those who did not use cannabis. These findings were partially accounted for by substance use*gender interactions: men using cannabis reported more vigorous activity than all other groups, and women with alcohol use reported less walking than men with and without alcohol use. Research is needed to increase physical activity among women who use substances and to evaluate reasons for the relationship between substance use and physical activity among men.
191
Allergic bronchopulmonary aspergillosis in association with rheumatic heart disease: report of three cases
Allergic bronchopulmonary aspergillosis is usually seen in patients with asthma or cystic fibrosis. Its association with rheumatic heart disease has not been adequately reported in literature. We report our experience of three cases who were diagnosed cases of rheumatic heart disease. Their symptomatology and clinical findings required further evaluation and investigations, which were suggestive of allergic bronchopulmonary aspergillosis. The patients were treated with steroids and/or antifungals before proceeding with the valve replacement.
192
A Study on the Public Landscape Order of Xinye Village
In the modernization process since China's reform and liberalization, urban and village space design is reflected in the characteristics of Western cultures. The idea of Western space design has a profound influence on China, but the piecemeal individuation of art design, the disorderly public art modeling and concept, not only interferes with the aesthetic sense of urban and village public space itself, but also seriously affected the landscape order of public space. In fact, Chinese traditional settlement landscape excels in abundant landscape design and spatial sequence. This paper, using the methods of literature discussion, field research and spatial analysis, takes the typical traditional landscape settlement Xinye Village (???) in the south of the Yangtze River as an example, and explores its public landscape order as a whole, and finds its spatial structure based on the Five Elements and Nine Divisions (????) cultural schemata. In the process of development, it has experienced the competition of public space, thus forming a stable and sustainable spatial order form. The purpose is to explore the cultural schema of the public landscape from the traditional Chinese settlement, and to put forward the possibility of constructing the public landscape order based on culture in future urban and village landscape design.
193
Feasibility of Delivering High-Dose Methotrexate in Adolescent and Adult All Patients: A Retrospective Study
Introduction HD-MTX is a key drug in the treatment protocols for ALL. The regimen needs to be administered with appropriate supportive measures and serum methotrexate level monitoring. A limited testing strategy is relevant in resource constraint settings since it allows a shorter duration of hospitalization. We report our experience with this strategy and its impact on the patient safety outcomes. Methods This is a retrospective study of all patients ≥ 15 years of age with newly diagnosed ALL or Lymphoblastic lymphoma (LBL) who were administered HDMTX (part of BFM-90 ALL protocol) at our institute between March 2013 to November 2013.The medical records were reviewed for clinical characteristics, disease-related details, HDMTX dose and cycles administered, leucovorin rescue and toxicities. Results A total of 423 cycles of HD-MTX were administered to 106 patients during the study period. The median duration for completion of all 4 cycles of HDMTX was 53 (IQR 49-60) days. The grade 3 or higher toxicities were anemia in 9.6%, neutropenia 19.4%, febrile neutropenia 5.7%, thrombocytopenia 4.4% and mucositis in 0.7%. There was statistically significant correlation between the levels at 42 h (≤ 1 mmol/L vs > 1 mmol/L) and toxicity- anemia, FN and mucositis observed more in the late clearance group. With limited sampling strategy whereby if the 42- hour level MTX level are < 1 mmol/L, 57% of patients could be discharged early. Conclusion HD-MTX can be safely administered to adolescent and adult ALL patients. A limited methotrexate level monitoring is a safe strategy that can optimize the resources better.
194
Monkeypox after Occupational Needlestick Injury from Pustule
We report a case of monkeypox in a physician after an occupational needlestick injury from a pustule. This case highlights risk for occupational transmission and manifestations of the disease after percutaneous transmission: a short incubation period, followed by a solitary lesion at the injured site and later by systemic symptoms.
195
Editorial: Whither globalization and health in an era of geopolitical uncertainty?
Globalization has been declared dead or dying for many years, although recently, the number of voices declaring it 'over' has swelled [1]. As editors of a journal interrogating how globalization affects health, we confront the question: Have the COVID-19 pandemic, Russia's war against Ukraine, a breakdown in multilateralism, and the risk of a return to the stagflation of the 1970s finally sounded a death knell for the research and scholarship we have been publishing in the journal's 20-year history? We think not and argue below why, in our post-pandemic fractured and fractious era, it is vitally important to retain a focus on this messy construct short-handed as 'globalization.'
196
Learning Structured Models for Segmentation of 2-D and 3-D Imagery
Efficient and accurate segmentation of cellular structures in microscopic data is an essential task in medical imaging. Many state-of-the-art approaches to image segmentation use structured models whose parameters must be carefully chosen for optimal performance. A popular choice is to learn them using a large-margin framework and more specifically structured support vector machines (SSVM). Although SSVMs are appealing, they suffer from certain limitations. First, they are restricted in practice to linear kernels because the more powerful nonlinear kernels cause the learning to become prohibitively expensive. Second, they require iteratively finding the most violated constraints, which is often intractable for the loopy graphical models used in image segmentation. This requires approximation that can lead to reduced quality of learning. In this paper, we propose three novel techniques to overcome these limitations. We first introduce a method to "kernelize" the features so that a linear SSVM framework can leverage the power of nonlinear kernels without incurring much additional computational cost. Moreover, we employ a working set of constraints to increase the reliability of approximate subgradient methods and introduce a new way to select a suitable step size at each iteration. We demonstrate the strength of our approach on both 2-D and 3-D electron microscopic (EM) image data and show consistent performance improvement over state-of-the-art approaches.
197
The Low Area Probing Detector as a Countermeasure Against Invasive Attacks
Microprobing allows intercepting data from on-chip wires as well as injecting faults into data or control lines. This makes it a commonly used attack technique against security-related semiconductors, such as smart card controllers. We present the low area probing detector (LAPD) as an efficient approach to detect microprobing. It compares delay differences between symmetric lines such as bus lines to detect timing asymmetries introduced by the capacitive load of a probe. Compared with state-of-the-art microprobing countermeasures from industry, such as shields or bus encryption, the area overhead is minimal and no delays are introduced; in contrast to probing detection schemes from academia, such as the probe attempt detector, no analog circuitry is needed. We show the Monte Carlo simulation results of mismatch variations as well as process, voltage, and temperature corners on a 65-nm technology and present a simple reliability optimization. Eventually, we show that the detection of state-of-the-art commercial microprobes is possible even under extreme conditions and the margin with respect to false positives is sufficient.
198
A new joint CTC-attention-based speech recognition model with multi-level multi-head attention
A method called joint connectionist temporal classification (CTC)-attention-based speech recognition has recently received increasing focus and has achieved impressive performance. A hybrid end-to-end architecture that adds an extra CTC loss to the attention-based model could force extra restrictions on alignments. To explore better the end-to-end models, we propose improvements to the feature extraction and attention mechanism. First, we introduce a joint model trained with nonnegative matrix factorization (NMF)-based high-level features. Then, we put forward a hybrid attention mechanism by incorporating multi-head attentions and calculating attention scores over multi-level outputs. Experiments on TIMIT indicate that the new method achieves state-of-the-art performance with our best model. Experiments on WSJ show that our method exhibits a word error rate (WER) that is only 0.2% worse in absolute value than the best referenced method, which is trained on a much larger dataset, and it beats all present end-to-end methods. Further experiments on LibriSpeech show that our method is also comparable to the state-of-the-art end-to-end system in WER.
199
Multi-Level Optimization of an Ultra-Low Power BrainWave System for Non-Convulsive Seizure Detection
We present a systematic evaluation and optimization of a complex bio-medical signal processing application on the BrainWave prototype system, targeted towards ambulatory EEG monitoring within a tiny power budget of <1 mW. The considered BrainWave processor is completely programmable, while maintaining energy-efficiency by means of a Coarse-Grained Reconfigurable Array (CGRA). This is demonstrated through the mapping and evaluation of a state-of-the-art non-convulsive epileptic seizure detection algorithm, while ensuring real-time operation and seizure detection accuracy. Exploiting the CGRA leads to an energy reduction of 73.1%, compared to a highly tuned software implementation (SW-only). A total of 9 complex kernels were benchmarked on the CGRA, resulting in an average 4.7 x speedup and average 4.4 x energy savings over highly tuned SW-only implementations. The BrainWave processor is implemented in 28-nm FDSOI technology with 80 kB of Foundry-provided SRAM. By exploiting near-threshold computing for the logic and voltage-stacking to minimize on-chip voltage-conversion overhead, additional 15.2% and 19.5% energy savings are obtained, respectively. At the Minimum-Energy-Point (MEP) (223 mu W, 8 MHz) we report a measured state-of-the-art 90.6% system conversion efficiency, while executing the epileptic seizure detection in real-time.