Search

With the development of modern cities, urban rail transit has become an indispensable part of residents' travelling mode, and accurate prediction of urban rail transit passenger flow is particularly important. However, due to the non-linearity and non-stability of passenger flow, the low quality of big data and the lack of data make it more and more difficult to predict the passenger flow of urban rail transit. In this paper, we propose a unique model structure that ingeniously integrates Gated Recurrent Units (GRU), Convolutional Neural Networks (CNN) with fused residual network connections, and a self-attention mechanism. The design of this model is primarily to effectively handle complex and low-quality data, which is often prevalent in passenger flow prediction scenarios. To validate the effectiveness of the proposed model, we used data from the Beijing urban rail transit in 2016 for prediction. In the experiments, we performed comparison study and ablation study. The experimental results show that the model proposed in this paper has significantly improved prediction accuracy in both horizontal and vertical comparisons. This outcome substantiates that the proposed model can not only effectively handle complex and low-quality data but also extract short-term features and long-term dependency features of passenger flow well, thereby achieving more accurate predictions. • Integration of GRU, CNN with fused residual connections, and self-attention mechanism constitutes a novel model. • The model is capable of handling complex and lower-quality data, significantly improving the accuracy of predictions. • The model can simultaneously extract short-term features and long-term dependency features from time series. [ABSTRACT FROM AUTHOR]

Traffic Anomaly Detection (TAD) is an important and difficult task in Intelligent Transportation Systems (ITS). Traffic anomaly events are sparse in both spatial and temporal spaces, posing a challenge to the performance of model. Moreover, a single traffic anomaly event can impact multiple road sections in the neighborhood, further undermining the accuracy of TAD. In this paper, we propose a new TAD method based on spatio-temporal hypergraph convolutional neural network. Specifically, we adopt a spatial–temporal augmentation approach for traffic data. This will enhance the performance of detecting sparse anomalies. Meanwhile, we introduce a hypergraph learning method to model the road network. This could capture the spreading features of anomalies for better detection results. Additionally, we design a dynamic hypergraph construction method to extract the evolving relationships of road segments. The proposed model evaluation on the Beijing (SE-BJ) dataset for TAD reveals superior performance compared to state-of-the-art ones. • Proposed a data augmentation method to enhance the detect effect. • Introduced hypergraph representation for effective road network modeling. • Designed a dynamic hypergraph representation to capture evolving road structures. • Proposed a spatio-temporal hypergraph convolutional network for TAD. [ABSTRACT FROM AUTHOR]

Pedestrian evacuation dynamics research is crucial for safety engineering and crowd management. Accurate and fast extraction of pedestrian trajectories from experimental videos is essential for reliable evacuation data and effective strategy development. In this paper, we propose a novel method for extracting pedestrian trajectories from evacuation experiment videos based on deep learning techniques. The method consists of two modules: pedestrian detection and location prediction, which use Convolutional Neural Networks (CNNs) and Long Short-Term Memory (LSTM) networks, respectively. Experimental results show that our method achieves high accuracy with short time compared to traditional methods. Furthermore, our experiments demonstrate that our method can accurately extract information on pedestrian evacuation dynamics. The proposed method provides a fast and reliable approach to extracting pedestrian evacuation dynamics information, which has the potential to be utilized by researchers, urban planners, and emergency management personnel to develop more effective evacuation strategies, improve crowd management, and ultimately enhance public safety. [ABSTRACT FROM AUTHOR]

This paper proposes a method for color-texture analysis by learning spatio-spectral representations from a complex network framework using the Randomized Neural Network (RNN). We model the color-texture image as a directed complex network based on the Spatio-Spectral Network (SSN) model, which considers within-channel connections in its topology to represent the spatial characteristics and spectral patterns covered by between-channel links. The insight behind the method is that complex topological features from the SSN can be embedded by a simple and fast neural network model for color-texture classification. Thus, we investigate how to effectively use the RNN to analyze and represent the spatial and spectral patterns from the SSN. We use the SSN vertex measurements to train the RNN to predict the dynamics of the complex network evolution and adopt the learned weights of the output layer as descriptors. Classification experiments in four datasets show the proposed method produces a very discriminative representation. The results demonstrate that our method obtains accuracies higher than several literature techniques, including deep convolutional neural networks. The proposed method also showed to be promising for plant species recognition, achieving high accuracies in this task. This performance indicates that the proposed approach can be employed successfully in computer vision applications. • A complex network spatio-spectral representation for color-texture classification is proposed. • It is based on combining a complex network model and randomized neural networks. • Color textures are modeled as spatio-spectral complex networks. • Randomized neural networks learn representations from the modeled complex networks. • The spatio-spectral representations are the output weights of the neural network. [ABSTRACT FROM AUTHOR]

Accurate traffic flow prediction is important for congestion identification and traffic dispersion. The original traffic flow data may generate different noises in the detector collection process and data aggregation process, resulting in large errors in the prediction results. To solve the problems above, this paper proposes a convolutional neural network model based on wavelet reconstruction (WT-2DCNN), which eliminates potential outliers in the data by introducing wavelet method, and constructs a WT-2DCNN model based on data extension and fusion of convolutional neural networks, and obtains the internal trend features of traffic flow through multiple convolutional and pooling layers in this model for traffic flow prediction. In this paper, the performance and training efficiency of the WT-2DCNN model has been validated on the publicly available Caltrans Performance Measurement System (PeMS) dataset, and the Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE) are significantly lower than those of other algorithms, and the training time is only 1/4 to 1/3 of that of Recurrent Neural Networks (RNN), indicating that the WT-2DCNN model has higher accuracy and more efficient training efficiency. • Prediction using only a single time-series data without timestamp and spatial data. • Fusing multiple wavelet reconstruction data for data extension. • Extracting deep features of traffic flow using two-dimensional convolutional neural networks. • Improving prediction accuracy and reducing training time. [ABSTRACT FROM AUTHOR]

Music is often considered as the language of emotions. The way it stimulates the emotional appraisal across people from different communities, culture and demographics has long been known and hence categorizing on the basis of emotions is indeed an intriguing basic research area. Indian Classical Music (ICM) is famous for its ambiguous nature, i.e. its ability to evoke a number of mixed emotions through only a single musical narration, and hence classifying evoked emotions from ICM becomes a more challenging task. With the rapid advancements in the field of Deep Learning, this Music Emotion Recognition (MER) task is becoming more and more relevant and robust, hence can be applied to one of the most challenging test case i.e. classifying emotions elicited from ICM. In this paper we present a new dataset called JUMusEmoDB which presently has 1600 audio clips (approximately 30 s each) where 400 clips each correspond to happy, sad, calm and anxiety emotional scales. The initial annotations and emotional classification of the database was done based on an emotional rating test (5-point Likert scale) performed by 100 participants. The clips have been taken from different conventional ' raga' renditions played in two Indian stringed instruments – sitar and sarod by eminent maestros of ICM and digitized in 44.1 kHz frequency. The ragas , which are unique to ICM, are described as musical structures capable of inducing different moods or emotions. For supervised classification purposes, we have used Convolutional Neural Network (CNN) based architectures (resnet50, mobilenet v2.0, squeezenet v1.0 and a proposed ODE-Net) on corresponding music spectrograms of the 6400 sub-clips (where every clip was segmented into 4 sub-clips) which contain both time as well as frequency domain information. Along with emotion classification, instrument classification based response was also attempted on the same dataset using the CNN based architectures. In this context, a nonlinear technique, Multifractal Detrended Fluctuation Analysis (MFDFA) was also applied on the musical clips to classify them on the basis of complexity values extracted from the method. The initial classification accuracy obtained from the applied methods are quite inspiring and have been corroborated with ANOVA results to determine the statistical significance. This type of CNN based classification algorithm using a rich corpus of Indian Classical Music is unique even in the global perspective and can be replicated in other modalities of music also. The link to this newly developed dataset has been provided in the dataset description section of the paper. This dataset is still under development and we plan to include more data containing other emotional as well as instrumental entities into consideration. • A novel database of 1600 emotional clips taken from Indian Classical music (ICM). • Application of nonlinear and CNN based methods to classify emotions and instruments from ICM. • Proposal of a new ODE-Net based architecture which provides the best classification rate. • Multifractal technique reflects the inherent ambiguity in emotions of ICM. [ABSTRACT FROM AUTHOR]

The research on theoretical techniques and practical applications of vehicle speed prediction holds significant importance in mitigating the increasingly pressing traffic issues and enhancing the operational efficiency of urban transportation systems. This paper aims to improve the accuracy of traffic speed prediction. To achieve this, we have devised a multi-graph convolutional neural network (GCN) framework that utilizes data from three different patterns: recent, daily, and weekly, enabling the extraction of distinct time-related features. Through the integration of multiple GCNs and 3D convolutional neural networks (3DCNN), we have obtained richer spatiotemporal feature representations. Additionally, we have applied wavelet analysis to process the traffic speed data from each node in the road network, unearthing hidden information and eliminating noise interference. Our experiments have demonstrated that the model employing wavelet analysis outperforms the baseline model without wavelet analysis, thus enhancing the prediction accuracy. In conclusion, we propose a wavelet-based spatiotemporal multi-graph convolutional neural network for traffic speed prediction. Our experimental results validate the superiority of this model across different time intervals and real-world datasets, showcasing its advantages in terms of prediction accuracy and robustness. As a result, our approach offers an effective and viable solution for traffic speed prediction tasks. • A multi-graph convolutional neural network predicts data jointly using recent, daily, and weekly modes, extracting diverse temporal features with GCNs and 3DCNNs. • Wavelet analysis on traffic speed data reveals hidden insights and improves the model compared to the baseline. • The spatiotemporal multi-graph convolutional neural network with wavelet analysis achieves superior accuracy over popular models in extensive experiments. [ABSTRACT FROM AUTHOR]

With the development of the internet and big data, the missing or hidden information identification of text data has become an imperative task. At present, the challenge in the hidden information study is judging whether there is hidden information and where it exists. In this paper, hidden information refers to the words that do not appear in a sentence, however, they have certain correlations with the existing words or sentence and have a great influence on the comprehension of a sentence or part of the text data. This paper focuses on discovering the key and influential hidden information in the text data. A keyword-based hidden information extraction framework is proposed in this paper to search hidden entities, with the assumption that the importance of hidden objects is reflected by the keywords in the text data. A network-based Convolution Neural Network (CNN) model is developed to identify the hidden information related to keywords. The model is based on the results of CNN, and cosine similarity is used to judge whether there is hidden information in the source text data or not. We primarily form the word co-occurrence network of text, select the words with the highest degree as keywords, and generate random walk paths on the network. Besides, we use the random walk path where the last word is the keyword to train CNN. In the experimental section, the proposed model is applied to the dataset in 20Newgroups. The results show that the proposed model can effectively identify the hidden information associated with the keywords in the source text data, and the detection accuracy of keywords can reach 98%–99% achieved by CNN. [ABSTRACT FROM AUTHOR]

Estimating the impacts of traffic control strategies (TCSs) can provide feedback in traffic control and help to identify the effective ones among massive strategies, thus boosting intelligent railway systems. This paper aims to assess the impacts of TCSs in railway systems from two perspectives: given a TCS on the target trains, for the subsequent trains, (1) how many control actions are needed in the future? (2) how many changes are there in terms of the average train delay? To this end, a hybrid learning model combining a convolutional neural network (CNN) and the random forest (RF), named CNN-RF, was innovatively proposed. The CNN component extracts features from the data, while the RF component predicts the targets (i.e., the effect measurements), to ensure the model's performance. The proposed model was evaluated based on the real-world train operation data from Chinese high-speed railways, and the average values of metrics f1-score, g-means, MAE (mean absolute error), RMSE (root mean square error), and R 2 (goodness of fit) reach 0.756, 0.762, 0.488 min, 1.035 min, and 0.795, respectively. Compared with the benchmarks, the proposed model improves the above metrics by 10.52% on average. These results demonstrate that the proposed model effectively predicts the impacts of TCSs in the near future, facilitating rail traffic control and potentially improving the quality of transportation services. • The impacts of traffic control strategies are measured from different perspectives. • A hybrid machine learning model is proposed to forecast the impacts. • The proposed model is superior to the standard models. • The model can be used to provide feedback for rail traffic control. [ABSTRACT FROM AUTHOR]

The emergence of cooperation between competing agents has been commonly studied through evolutionary games, but such cooperation often requires a mechanism or a third party to be activated and kept alive. To investigate how a mechanism affects the evolution of cooperation, this paper proposes an innovative reinforcement learning-based strategy updating model. The model consists of two symmetrical sets of convolutional neural networks. Besides, the agents' strategies updating rules are defined: firstly, the agents learn and predict the environment and the behaviors of neighboring agents, then estimate their future payoffs based on this information, and finally determine their strategies based on these estimated payoffs. Through investigating the behavior characteristics and the stable states of the network for highly intelligent agents with memory learning and prediction ability in the evolution of the prisoner's dilemma game, the results demonstrate that the game initiators who adopt the mixed optimal payoff approach can increase the number of cooperators and facilitate "global cooperation" and "repaying kindness with kindness". Although the temptation factor has little effect on the population, increasing the discount factor can expand the scale of the cooperative cluster and even achieve dynamic stability. Additionally, a smaller size of minibatch is beneficial for the evolution of cooperation in a smaller experience replay pool. A larger size of minibatch is more conducive to the evolution of cooperation with an increasing capacity of the experience replay pool. This research provides a novel perspective from reinforcement learning to understand the evolution of cooperation. • An innovative RLSUM is proposed to investigate the PDG. • The behavioral characteristics and the steady state of the network are studied. • The RLSUM can facilitate "global cooperation" and "repaying kindness with kindness". • The research provides a novel perspective from RL to understand the cooperative evolution. [ABSTRACT FROM AUTHOR]

Community detection is a fundamental problem for many networks, and many methods have been proposed to resolve it. However, due to rapid increases in the scale and diversity of networks, the modular organization at the global level in many large networks is often extremely difficult to recognize. In this paper, we propose a new method based on deep learning on ground-truth communities, with the aim of revealing community structure in large real-world networks. The contributions of this paper are 1) proposing an edge-to-image (E2I) model that can transfer the edge structure to an image structure; 2) construction of a community network (ComNet) to classify the two types of edges, which are those in the same community and others between different communities; 3) making it easier to obtain local views of network communities by breadth-first search based on edge classification; and 4) merging preliminary communities with local modularity R, making it easy to optimize the community structure and obtain the final community structure of given networks. The experimental results show that the proposed edge classification method based on deep convolution neural networks can increase the accuracy of community structure evaluation compared with the existing methods in computer-generated networks and large-scale real-world networks. • Proposing an edge-to-image (E2I) model that can transfer the edge structure to an image structure. • Construction of CNN(ComNet) to classify the two types of edges which are those in the same community and others between different communities. • The community structure of a network can be divided by removing the edges between different communities. • Use the local modularity R to merge communities to optimize the community structure. • The experimental results have better performance and accuracy in large scale network. [ABSTRACT FROM AUTHOR]

Recognition of abnormal driving behavior is an important application area as it can support driving reliability and improve safety. In the last decade, deep learning methods have been presented through fruitful academic research and industrial applications. State-of-the-art deep learning methods are not commonly used for detection of abnormal driving behavior based on driving parameter information, and are lacking in terms of recognition accuracy. Based on this, a novel data-driven abnormal driving behaviors method is proposed in this paper by combining a convolutional neural network (CNN) and a Bidirectional gated recurrent unit (BiGRU). In this process, real vehicle driving data, including the extreme acceleration and steering position, are analyzed to establish a dataset of driving behaviors recognition firstly. Then, the datasets are inputted into the CNN-BiGRU algorithm to recognize the abnormal driving behavior where CNN captures non-linear relations from long-term trends of sequences and BiGRU extracts features of time series from driving parameters. The experimental results show that the proposed method offers improved accuracy and robustness in recognizing abnormal driving compared with other existing machine learning methods. [ABSTRACT FROM AUTHOR]

With the development of society, the number of old people is increasing. Slow response, osteoporosis and vision loss threaten the health of the elderly. The fall of this problem is an important factor that threatens the health of the elderly. In order to reduce the damage caused by falls, this paper based on the human skeleton map for fall prediction. First using OPENPOSE get the bone map and make it into a data set. Then using transfer learning to train the data set to get a new model Finally, the new model is used to predict the fall. The innovations in this paper are to take bone maps from 2D images and use bone maps to make fall predictions. The bone map is predicted using a convolutional neural network. The final experimental results show that the new model obtained through transfer learning has an accuracy rate of 91.7%. This result fully demonstrates the validity of the proposed model. • The innovations in this paper are to obtain bone maps from 2D images. • This paper uses convolutional neural networks for fall prediction. • This paper uses OPENPOSE to get the bone map and make data sets. • This paper uses transfer learning to train the data set. [ABSTRACT FROM AUTHOR]

Real-time and accurate traffic state identification can provide reference for urban traffic control and guidance. Due to the randomness and complexity of traffic flow, it is difficult to identify the traffic state accurately. The existing researches mainly adopted traffic state feature vectors and machine learning to identify the traffic state. Few studies attempted to use the feature map and deep learning for traffic state recognition. Therefore, this paper proposes a traffic state recognition model based on the traffic state feature map and deep learning. The feature map is a chromatogram form of digital traffic state feature vector. And the deep learning has strong predictive performance in image identification. In the model, the road traffic state feature vector is extracted from the vehicle trajectory data, and the Gram Angle Field (GAF) is adopted to transform the feature vector into feature map. Then, the deep learning algorithm is utilized in traffic state offline classification and online recognition. In the offline training phase, the historical vehicle trajectory data and the DeepCluster algorithm are used to establish the mapping relationship between the feature maps and the traffic states. In the online identification stage, the real-time vehicle trajectory data and CoAtNet algorithm are utilized to identify the current traffic state. In experiments, the proposed model is compared with the Convolutional Neural Network (CNN), Support Vector Machine (SVM), and K-Nearest Neighbor (KNN) models using the open vehicle trajectory data of Shanghai North Cross Channel. The results showed that the proposed model achieved the best performance with the accuracy of 92.06% and could provide support for road traffic control and guidance. • Proposing a traffic state online recognition framework based on traffic state feature map and deep learning. • Adopting Gram Angle Field to construct traffic state feature map. • Using vehicles' trajectory data and CoAtNet to identify the traffic state. [ABSTRACT FROM AUTHOR]

Improving the traffic efficiency of the existing roads in the city has become an important task for the traffic management department. Timely and accurate traffic prediction is the key to urban traffic control and guidance. The multi-layer automatic feature extraction and expression ability of deep learning is deeply integrated with the non-linearity, multi-modality, and spatiotemporal correlation of traffic flow. A multi-step ahead traffic speed prediction based on gated temporal graph convolution network (GT-GCN) is proposed in this paper. Firstly, a dynamic weighted graph network is constructed according to the traffic road network structure, and then a graph convolutional network is used to process the graph structure data of the traffic road network and capture the spatial characteristics of the traffic flow. Subsequently, a gated temporal convolutional network is applied to capture the short- and long-range temporal correlations of traffic flow. Finally, a multi-step ahead prediction model based on a hybrid deep learning framework GT-GCN is proposed. Experiments using real datasets show that the proposed GT-GCN has a significant multi-step predictive performance compared to the baseline models. • A weighted dynamic graph network is constructed to characterize the topology of the transportation network. • A multi-step traffic prediction GT-GCN is proposed based on graph convolutional network and gated temporal convolutional network. • Extensive experiments on real-world traffic show that the proposed GT-GCN achieves significant predictive performance. [ABSTRACT FROM AUTHOR]

Convolutional neural networks have been shown to extract features better than traditional algorithms in the fields such as image classification, object detection, and speech recognition. In parallel, a variational quantum algorithm incorporating parameterized quantum circuits has higher performance on near-term quantum processors. In this paper, we propose a classification algorithm called variational convolutional neural networks (VCNN), allowing for efficient training and implementation on near-term quantum devices. The VCNN algorithm combines the multi-scale entanglement renormalization ansatz. We deploy the VCNN algorithm on the TensorFlow Quantum platform with the numerical simulator backends using the MNIST and Fashion MNIST datasets. Experimental results show that the average accuracy of VCNN on classification tasks can reach up to 96.41%. Our algorithm has higher learning accuracy and fewer training epochs than quantum neural network algorithms. Moreover, we conclude that circuit-based models have excellent resilience to noise by numerical simulations. • Allowing for efficient training and implementation on near-term quantum devices. • Combines the multi-scale entanglement renormalization ansatz. • The average accuracy of variational convolutional neural networks on classification tasks can reach up to 96.41%. • Higher learning accuracy and fewer training epochs. • Found circuit-based models have excellent resilience to noise. [ABSTRACT FROM AUTHOR]

Prediction of passenger flow distribution in urban rail transit stations can provide important data support for passenger flow organization and passenger travel guidance. However, complex station space structure and simulation-based passenger flow data bring challenges to accurate analysis and prediction of the passenger flow inside the station. This paper proposes a temporal graph attention convolutional neural network model (TGACN) to predict the passenger flow volume and density in key areas of the station. Firstly, considering the topological structure of key areas and the characteristics of passenger flow and flow trend in the station, a multi-graph generation method for continuous space in stations is designed, including geographic neighborhood graph and semantic neighborhood graph, to represent the static and dynamic correlation between nodes. Secondly, a new method of spatio-temporal feature fusion is proposed, which takes multi-graph as input to optimize the extraction and expression of spatial and temporal correlation. Finally, the TGACN is verified by passenger flow data set, which is constructed based on real-time video monitoring data of a transit station in Guangzhou. Experiments demonstrate that the TGACN can obtain the spatio-temporal correlation from passenger flow data, and the prediction results are better than the existing baseline models. • A prediction model for multipoint passenger flow in station is proposed. • Geographical graph is constructed to measure the spatial correlation between nodes. • Semantic graph is constructed to measure the correlation of passenger flow state. • Effective spatial information screening is realized by dimensionality reduction. [ABSTRACT FROM AUTHOR]

Vehicle trajectory prediction is essential for the operation safety and control efficiency of automated driving. Prevailing studies predict car following and lane change processes in a separate manner, ignoring the dependencies of these two behaviors. To remedy this issue, this paper proposes an integrated deep learning-based two-dimension trajectory prediction model that can predict combined behaviors. Specifically, we designed a switch neural network structure based on the attention mechanism, bi-directional long-short term memory (BiLSTM) and Temporal convolution neural network (TCN) to mimic and predict the joint behaviors. Experiments are conducted based on the Next Generation Simulation (NGSIM) dataset to validate the effectiveness of our proposed model. As results indicate, our proposed model outperforms the state-of-art trajectory prediction models and can provide accurate short-term and long-term predictions. • Integrated Car-following and Lane Change Prediction Algorithm. • A Deep Neural Network with a Switch Structure for the Integrated Behavior Prediction. • A Deep Neural Network with a Temporal Convolution Network and Attention Mechanism for Prediction Accuracy Improvement. [ABSTRACT FROM AUTHOR]

Real-time and accurate traffic state forecasting of urban roads is of great significance to improve traffic efficiency and optimize travel routes. However, future traffic state forecasting is still a challenging issue as it is influenced by several complicated factors including the dynamic spatio-temporal dependencies. Existing models usually consider the dependencies from the road sections with physical connections and ignore the road sections without physical connections. To this end, this paper proposes a deep ensemble neural network (DENN) model to improve the accuracy of urban traffic state forecasting by forming the road sections with high relevance into a virtual graph. To capture the spatio-temporal characteristics efficiently and simultaneously, the DENN integrates the graph convolutional neural network, bidirectional gated recurrent unit network, and a dense layer with attention mechanism into an end-to-end fashion. Validated on two ground-truth urban traffic speed datasets, the DENN model can well fit the nonlinear fluctuation of urban speed and indicate superior performance than the state-of-the-art benchmark methods in terms of prediction precision and robustness. • The virtual network is established by forming the road sections with high relevance into a virtual graph. • The graph convolutional network (GCN) is introduced to mine spatial features of the traffic flow from virtual graph. • Deep ensemble neural network is built by fusing a GCN, Bi-GRU network, and attention model into an end-to-end fashion. • Real-world urban traffic datasets are used to verify the proposed model in terms of prediction accuracy and stability. [ABSTRACT FROM AUTHOR]

Traffic prediction plays an essential role in the intelligent transportation systems and has broad applications in transportation management and planning. And the key to this field is to explore the spatiotemporal information of traffic data, synchronously. Recently, various deep learning methods, such as Convolution Neural Network (CNN) and Graph Convolutional Network (GCN), have shown promising performance in traffic prediction. However, these methods cannot automatically model spatial dependencies and dynamic spatiotemporal States, and there are no constraints on the distribution of outputs. To solve the above problems, in this paper, a method of automatically obtaining spatiotemporal dependence in data, which can automatically obtain the spatiotemporal state and spatiotemporal dependency using Multi-graph Adversarial Neural Network (GAN) and is named AST-MAGCN, is proposed. The new method AST-MAGCN combines GAN and GCN, extracts spatiotemporal state of the data in real-time, and outputs the traffic forecast by GAN constraint. Lastly, the proposed method is evaluated on two real-world traffic datasets, and the experimental results show that the proposed method outperforms baseline traffic prediction methods. [ABSTRACT FROM AUTHOR]

Epilepsy is one of the most common brain diseases, and seizures usually occur randomly. Accurately predicting seizures enable doctors and patients to carry out medical prevention timely. In seizure prediction studies, single-domain information input (time domain or frequency domain, Etc.) neglects some parts' information from signals. In this paper, we propose a novel deep learning framework named channel attention dual-input convolutional neural network (CADCNN) to obtain the signal's useful information fully. The spatial–temporal features extracted by short-time Fourier transform (STFT) are fed to the CADCNN, and the raw EEG signals are fed for further feature extraction. With the fusion of two inputs from different domains and the combination of channel attention, CADCNN can learn faithful and distinguishable representations of EEG signals and boost the temporal, spectrum, and spatial information utilization capability. We evaluate the proposed method using the Boston Children's Hospital-MIT scalp EEG public datasets. Compared with other state-of-the-art methods, the sensitivity, false prediction rate, specificity, and AUC of our proposed method reach 97.1%, 0.029h, 95.6%, and 0.917, respectively, presenting better performance and higher prediction accuracy. • We develop channel attention dual-input convolutional neural network to analyze epileptic EEG signals. • Our method enhances the model's ability to capture Spatio-temporal information. • Our method achieves high sensitivity and low false prediction rate. [ABSTRACT FROM AUTHOR]

In gene expression, high-resolution Drosophila embryonic images contain abundant temporal and spatial information. The Drosophila embryo of interest detection, with high accuracy and rapidity, is an important preprocessing step in the Drosophila embryonic gene expression computation system. In this paper, we proposed a novel multi-feature fusion (MFF) CNNs framework for the Drosophila embryo of interest detection. Considering the great variety of Drosophila embryonic images, the proposed network takes full advantages of multi-level and multi-scale convolutional features by leveraging the deeply-supervised nets and side-output layers. We built a Drosophila Embryonic Dataset, and train our framework with the Dataset. In the experiment, our method yielded satisfactory results, with advantages in terms of high accuracy (94.9% mean F-measure) and efficiency (40 FPS, i.e. Frame per Second). To the best of our knowledge, it is the first attempt to solve this problem with CNNs and achieves good results. • The proposed framework exploits all the meaningful convolutional features to represent the complex morphology and gene expression patterns. • We constructed a Drosophila Embryo DataSet (DEDS). • We have evaluated the proposed model with F-measure, Area Under Curve and Mean Absolute Error. [ABSTRACT FROM AUTHOR]