Search

Intelligent Traffic Light Control System (ITLCS) is a typical Multi-Agent System (MAS), which comprises multiple roads and traffic lights.Constructing a model of MAS for ITLCS is the basis to alleviate traffic congestion. Existing approaches of MAS are largely based on Multi-Agent Deep Reinforcement Learning (MADRL). Although the Deep Neural Network (DNN) of MABRL is effective, the training time is long, and the parameters are difficult to trace. Recently, Broad Learning Systems (BLS) provided a selective way for learning in the deep neural networks by a flat network. Moreover, Broad Reinforcement Learning (BRL) extends BLS in Single Agent Deep Reinforcement Learning (SADRL) problem with promising results. However, BRL does not focus on the intricate structures and interaction of agents. Motivated by the feature of MADRL and the issue of BRL, we propose a Multi-Agent Broad Reinforcement Learning (MABRL) framework to explore the function of BLS in MAS. Firstly, unlike most existing MADRL approaches, which use a series of deep neural networks structures, we model each agent with broad networks. Then, we introduce a dynamic self-cycling interaction mechanism to confirm the "3W" information: When to interact, Which agents need to consider, What information to transmit. Finally, we do the experiments based on the intelligent traffic light control scenario. We compare the MABRL approach with six different approaches, and experimental results on three datasets verify the effectiveness of MABRL.

Traffic intersections are important scenes that can be seen almost everywhere in the traffic system. Currently, most simulation methods perform well at highways and urban traffic networks. In intersection scenarios, the challenge lies in the lack of clearly defined lanes, where agents with various motion plannings converge in the central area from different directions. Traditional model-based methods are difficult to drive agents to move realistically at intersections without enough predefined lanes, while data-driven methods often require a large amount of high-quality input data. Simultaneously, tedious parameter tuning is inevitable involved to obtain the desired simulation results. In this paper, we present a novel adaptive and planning-aware hybrid-driven method (TraInterSim) to simulate traffic intersection scenarios. Our hybrid-driven method combines an optimization-based data-driven scheme with a velocity continuity model. It guides the agent's movements using real-world data and can generate those behaviors not present in the input data. Our optimization method fully considers velocity continuity, desired speed, direction guidance, and planning-aware collision avoidance. Agents can perceive others' motion planning and relative distance to avoid possible collisions. To preserve the individual flexibility of different agents, the parameters in our method are automatically adjusted during the simulation. TraInterSim can generate realistic behaviors of heterogeneous agents in different traffic intersection scenarios in interactive rates. Through extensive experiments as well as user studies, we validate the effectiveness and rationality of the proposed simulation method., Comment: 13 pages, 12 figures

Graph convolutional networks (GCNs) based methods have achieved advanced performance on skeleton-based action recognition task. However, the skeleton graph cannot fully represent the motion information contained in skeleton data. In addition, the topology of the skeleton graph in the GCN-based methods is manually set according to natural connections, and it is fixed for all samples, which cannot well adapt to different situations. In this work, we propose a novel dynamic hypergraph convolutional networks (DHGCN) for skeleton-based action recognition. DHGCN uses hypergraph to represent the skeleton structure to effectively exploit the motion information contained in human joints. Each joint in the skeleton hypergraph is dynamically assigned the corresponding weight according to its moving, and the hypergraph topology in our model can be dynamically adjusted to different samples according to the relationship between the joints. Experimental results demonstrate that the performance of our model achieves competitive performance on three datasets: Kinetics-Skeleton 400, NTU RGB+D 60, and NTU RGB+D 120., Comment: 12 pages, 6 figures

Additional file 1: Fig. S1. Distribution of codon pair bias scores in human, canine and mouse. CPBSs were calculated using the previous described method [17], Each salmon red circle represents a CPBS of a single human (a), canine (b), and mouse (c) gene plotted against its gene length, respectively. The average CPBSs of the human, canine and mouse are 0.0703, 0.0704, and 0.0651, respectively. Fig. S2. The sequence of the ooG. The ooG sequence was produced based on the oG [15] by using a previous described method [17]. > ooG sequence ATGGTTCCTCAAGCCCTTCTCTTTGTTCCTCTTCTTGTCTTCCCGCTCTGCTTTGGGAAGTTCCCCATCTACACCATTCCTGACAAGCTAGGGCCCTGGAGCCCCATTGACATCCACCACCTCAGCTGCCCCAACAACTTGGTTGTAGAAGATGAAGGCTGCACCAACCTCAGTGGCTTCTCCTACATGGAGCTAAAAGTGGGCTACATCTCGGCCATCAAGATGAATGGCTTCACCTGCACTGGAGTTGTCACTGAAGCAGAGACCTACACCAACTTTGTTGGCTATGTCACCACCACCTTCAAAAGAAAACACTTCCGGCCCACTCCAGATGCCTGCCGCGCGGCCTACAACTGGAAGATGGCGGGGGACCCCCGCTATGAAGAGAGCCTGCACAACCCCTACCCAGACTACCACTGGCTGAGGACTGTGAAGACCACCAAAGAAAGTTTGGTCATCATCAGCCCCAGTGTAGCTGACTTGGACCCCTATGACCGTTCTCTACACAGCCCTGTATTTCCTGGTGGGAACTGCAGTGGTGTGGCTGTCAGCAGCACCTACTGCAGCACCAACCATGACTACACCATCTGGATGCCGGAGAACCCCCGGCTAGGGATGTCCTGTGACATCTTCACCAACAGCCGAGGGAAAAGAGCCAGCAAAGGTTCTGAGACCTGTGGCTTTGTAGATGAGCGTGGCCTCTACAAGAGTTTAAAAGGTGCCTGCAAATTAAAACTCTGTGGTGTTCTTGGTCTTCGGCTCATGGATGGCACCTGGGTGGCCATGCAGACCAGCAATGAGACCAAGTGGTGCCCGCCGGGCCAGCTTGTCAACCTCCATGACTTCCGAAGTGATGAAATAGAACATCTTGTTGTAGAAGAACTTGTCAAGAAAAGAGAAGAATGTTTAGATGCCCTGGAGAGCATCATGACTACCAAGAGTGTCTCCTTCCGTCGCCTCAGCCACCTCAGGAAACTTGTTCCTGGCTTTGGGAAAGCCTACACCATCTTCAACAAGACGCTCATGGAAGCAGATGCCCACTACAAATCTGTCCGCACGTGGAATGAGATCATTCCTTCCAAAGGCTGCCTCCGAGTTGGTGGCCGCTGCCACCCACATGTCAATGGTGTCTTCTTCAATGGCATCATTCTTGGGCCAGATGGAAATGTCCTCATTCCAGAGATGCAGAGCAGCCTGCTGCAGCAGCACATGGAACTTCTTGTCAGCAGTGTCATCCCGCTCATGCACCCGCTGGCAGACCCCAGCACTGTCTTCAAGAATGGAGATGAAGCAGAAGATTTTGTAGAAGTTCATCTTCCTGATGTTCATGAAAGAATTTCTGGTGTGGACTTGGGTCTTCCCAACTGGGGAAAATATGTTCTTCTTTCTGCTGGGGCGCTCACGGCGCTCATGTTAATAATATTCCTCATGACCTGCTGcAGAAGAGTCAACCGCTCGGAGCCCACCCAGCACAACCTTCGTGGCACGGGCCGAGAAGTTTCTGTCACGCCGCAGAGTGGGAAGATCATCTCCTCCTGGGAGAGCCACAAGTCAGGAGGAGAGACGCGCCTGTAA. Fig. S3. The effects of the ooG on labeling specificity and cellular toxicity. (a) The percentage of input neurons from each site to total quantified inputs in the ooG and oG groups. (b) The signals of the caspase-3 was determined by immunohistochemistry using antibody against the caspase-3 in the ooG and oG groups. (b) The fine structure of the neuron is apparent in the ooG and oG groups. These images are the representatives from three mice.

The generation process of medical image will inevitably introduce certain noises. These noises will degrade the image quality and affect the final clinical diagnosis. Therefore, denoising plays an important role in the pre-processing of medical image before the formal diagnosis and treatment. In this paper, the classical NLM algorithm is improved to denoise medical images by involving a novel noise weighting function and parallelizing. In our experiment, plenty of medical images have been tested and experiment results show that our algorithm can achieve better results and higher efficiency compared with the original NLM method.

Free Fatty Acid 4 receptor (FFA4 receptor or GPR120), a rhodopsin-like G protein coupled receptor (GPCR) subfamily member, is a receptor that senses specific fatty acids such as ω-3 fatty acid in fish oil or the endogenous signaling lipid, PHASA. FFA4 receptor is enriched in lung, colon and adipose tissue but is also detected in many other tissues and cells. The activation of FFA4 receptor has multiple effects, including but not limited to inhibition of inflammation, improving insulin sensitivity and adipogenesis, and regulating hormone secretion from the gastro-intestinal system and pancreatic islets. The important role of FFA4 receptor in maintaining metabolic homeostasis strongly indicates the great potential of selective FFA4 receptor agonizts to treat diabetes and inflammation. In this review, we summarize recent research progress in the physiological and biochemical studies of FFA4 receptor and highlight its underlying signaling mechanisms and ligand identification to assist future research to exploit FFA4 receptor as a drug target.

Figure S1. Characterization and quantification of exosome levels and heatmap of the up-regulated exosomal lncRNAs and mRNAs identified in the screening. (A) Transmission electron microscopy (TEM) revealed the phenotype of exosomes isolated from stage I GC patient and healthy control plasma samples. Scale bar, 100 nm. (B) TEM revealed the phenotype of exosomes isolated from GCC and HPSEC culture media. Scale bar, 200 nm. (C and D) NanoSight particle tracking analysis of the size distributions and number of exosomes isolated from stage I GC patient and healthy control plasma samples (C) and isolated from GCC and HPSEC culture media (D). (E and F) Western blot showing the presence of specific exosome marker proteins (CD9 and CD63) and the absence of tubulin in exosomes derived from stage I GC patient and healthy control plasma samples (E) and derived from GCC and HPSEC culture media (F). (G) Heatmap of the 79 up-regulated exosomal lncRNAs and mRNAs between the plasma samples of stage I GC patients (n = 10) and healthy controls (n = 5). (H) Heatmap of the 285 up-regulated exosomal lncRNAs and mRNAs between GCC (n = 4) and HPSEC (n = 4) culture media. (PDF 395 kb)

Table S4. Performance of CEA, IncUEGC1 and IncUEGC2 levels in the classification of GC and Healthy subsets, Stage I and Healthy subsets, Stage I and CAG subsets. (PDF 154Â kb)

This paper proposes a method based on repulsive forces and sparse reconstruction for the detection and location of abnormal events in crowded scenes. In order to avoid the challenging problem of accurately tracking each specific individual in a dense or complex scene, we divide each frame of the surveillance video into a fixed number of grids and select a single representative point in each grid as the individual to track. The repulsive force model, which can accurately reflect interactive behaviors of crowds, is used to calculate the interactive forces between grid particles in crowded scenes and to construct a force flow matrix using these discrete forces from a fixed number of continuous frames. The force flow matrix, which contains spatial and temporal information, is adopted to train a group of visual dictionaries by sparse coding. To further improve the detection efficiency and avoid concept drift, we propose a fully unsupervised global and local dynamic updating algorithm, based on sparse reconstruction and a group of word pools. For anomaly location, since our method is based on a fixed grid, we can judge whether anomalies occur in a region intuitively according to the reconstruction error of the corresponding visual words. We experimentally verify the proposed method using the UMN dataset, the UCSD dataset and the Web dataset separately. The results indicate that our method can not only detect abnormal events accurately, but can also pinpoint the location of anomalies.

QR codes are usually scanned in different environments, so they must be robust to variations in illumination, scale, coverage, and camera angles. Aesthetic QR codes improve the visual quality, but subtle changes in their appearance may cause scanning failure. In this paper, a new method to generate scanning-robust aesthetic QR codes is proposed, which is based on a module-based scanning probability estimation model that can effectively balance the tradeoff between visual quality and scanning robustness. Our method locally adjusts the luminance of each module by estimating the probability of successful sampling. The approach adopts the hierarchical, coarse-to-fine strategy to enhance the visual quality of aesthetic QR codes, which sequentially generate the following three codes: a binary aesthetic QR code, a grayscale aesthetic QR code, and the final color aesthetic QR code. Our approach also can be used to create QR codes with different visual styles by adjusting some initialization parameters. User surveys and decoding experiments were adopted for evaluating our method compared with state-of-the-art algorithms, which indicates that the proposed approach has excellent performance in terms of both visual quality and scanning robustness., Comment: 15pages

In this paper we present a novel crowd simulation method by modeling the generation and contagion of panic emotion under multi-hazard circumstances. Specifically, we first classify hazards into different types (transient and persistent, concurrent and non-concurrent, static and dynamic ) based on their inherent characteristics. Then, we introduce the concept of perilous field for each hazard and further transform the critical level of the field to its invoked-panic emotion. After that, we propose an emotional contagion model to simulate the evolving process of panic emotion caused by multiple hazards in these situations. Finally, we introduce an Emotional Reciprocal Velocity Obstacles (ERVO) model to simulate the crowd behaviors by augmenting the traditional RVO model with emotional contagion, which combines the emotional impact and local avoidance together for the first time. Our experimental results show that this method can soundly generate realistic group behaviors as well as panic emotion dynamics in a crowd in multi-hazard environments., Comment: Already accepted by IEEE Transactions on Systems, Man, and Cybernetics 2019, 15 pages and 13 figures

Table S1. Clinical characteristics of patients and healthy individuals of all sets. (PDF 176Â kb)

Figure S2. qPCR analysis of the relative lncUEGC1 and lncUEGC2 levels in exosomes and exosomes-depleted plasma from stage I and II GC patients (n = 5). Differences with P

Methods. (PDF 289Â kb)

PurposeIt is not an easy and simple task to manufacture a complete denture with high quality. Traditionally, it often needs a medical expert with experience and hand‐on skill, due to the manual way of denture manufacturing. The purpose of this paper is to implement the multi‐manipulator tooth‐arrangement robot system that can fully automate the denture manufacturing process.Design/methodology/approachA novel complete denture manufacturing mechanism is designed, which is based on the multi‐manipulator and dental arch generator. The visual tooth‐arrangement and robot control software is developed in VC++6.0. Preliminary experiments on tooth‐arrangement have been conducted using the proposed multi‐manipulator tooth‐arrangement robot prototype system.FindingsThe multi‐manipulator tooth‐arrangement robot prototype system can automatically design and manufacture a set of complete denture that fit a patient by visual tooth‐arrangement and robot control software according to the patient's jaw arch parameters.Research limitations/implicationsThe implication of research is that it is feasible that the manufacture strategy of complete denture fulfilled by multi‐manipulator tooth‐arrangement robot. The limitation of research is that it is difficult to realize coordinate control.Originality/valueThe traditional manual method which makes complete denture by medical personal experience will be changed after the multi‐manipulator tooth‐arrangement robot system is manufactured, and adjustment to each tooth position and orientation will be realized by this system.