Search

Emotion cause pair extraction is a subtask in the sentiment analysis task, which aims to extract all emotion clauses in a given document and the cause clauses corresponding to the emotion. The previous work ignores the interrelationship between the emotion clause and the cause clause when generating the expression of the emotion clause and the cause clause. In order to solve the problem, based on the idea of graph-based dependency parsing and incorporating the graph attention mechanism, this paper proposed the GAT-ECPE model. When the model obtained the expression of the emotion clause and the reason clause, it used the sentence vector as a node into the graph attention layer to learn the information about the relationship between the clauses, and then performed biaffine transform to obtain the encoding of the emotion cause pair expression. And it set up a multi-task to establish a relationship between the extraction of emotions and causes. The experimental results on the ECPE data set prove that compared with the previous series of models, this model has improved evaluation indicators. [ABSTRACT FROM AUTHOR]

Aiming at the problem of low segmentation accuracy caused by ignoring its essence in real-time semantic segmentation methods, this paper proposed a multi-level context guided lightweight network. Firstly, it combined the deep-wise separable convolution with asymmetric convolution, and designed a context guided model based on the parallel asymmetric convolution to learn the joint features of local feature and its surrounding contexts. Furthermore, this research not only stacked the model in the network to achieve multi-level feature optimization, but also selected the shallow features consistent with higher-stage semantics by channel attention model, so as to improve the segmentation accuracy. According to the experimental results, the average intersection and union ratio of the proposed network on Cityscapes datasets is 72. 4 % at 94. 7 frame rate, and this network achieves significant performance improvement on CamVid datasets. In comparison to the other current real-time semantic segmentation methods, the performance of the proposed network is better. [ABSTRACT FROM AUTHOR]

In view of the slow detection speed of traditional grain flour species, this study implemented the random forest algorithm-assisted microwave nondestructive testing technology to identify grain flour species efficiently and accurately based on ZYNQ platform. By analyzing the hardware implementation of the random forest model, this paper proposed an improved model parameter structure, which could effectively save the consumption of hardware storage resources. In order to shorten the prediction time of the algorithm and reduce the power consumption of the system, it introduced the early termination recognition mechanism in the hardware implementation to avoid the unnecessary decision tree prediction process on the premise of keeping the accuracy unchanged. In view of the Zedboard development board, it designed a storage scheme for model parameters to make full use of on-board resources and ensure the normal operation of the system. The experimental results show that the implementation time of the scheme with ZYNQ is about 54. 2% shorter than that of the random forest algorithm with traditional CPU, and no loss of recognition accuracy is caused. [ABSTRACT FROM AUTHOR]

The existing multi-label learning algorithms tend to focus on the forward projection from the instance space to the label space, and the problem of the loss of instance space information due to the reduction of the feature dimension during forward projection is often ignored. Hence, this paper proposed a multi-label classification algorithm with bidirectional mapping learning. Firstly, the method used the forward mapping loss from instance space to label space to established a linear multi-label classification model. Secondly, the bidirectional mapping model based on reconstruction loss regularization compensated for the discriminatory information loss in the forward mapping process. Finally, the bidirectional mapping mod el combined label correlation and instance correlation to fully exploit the potential relationship between labels and instances, and improved the ability of the model to handle nonlinear data through nonlinear kernel mapping. The experimental results show that compared with several other methods in recent years, the average performance improvements of the method in terms of Hamming loss, one error and ranking loss are 17. 68%, 17. 01% and 18. 57%. Average performance improvement of 12. 37% on six evaluation metrics, which verifies the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]

With the rising of SDN and achieving great success of machine learning methods in classification, prediction and control tasks, to find new TE technology, manage or route the traffic in the network adaptively and dynamical, ensuring QoS and improving quality of user experience (QoE) have become the focus of network research. Firstly, this paper introduced the basic structure of SDN and the research content and objectives of SDN traffic engineering. Secondly, it analyzed the application of supervised learning methods and reinforcement learning in SDN traffic engineering, and analyzed the advantages and disadvantages of existing algorithms. Finally, it summarized the future research directions and challenges. [ABSTRACT FROM AUTHOR]

Aiming at the security problems caused by fraudulent accounts in the blockchain, this paper proposed a machine learning-based fraud account detection and feature analysis model. It took the real on-chain data on Ethereum into feature extraction and used it as the data source of the model. It obtained the optimal model by comparing different machine learning methods and performing iterative training to obtain the best predictive model. At the same time, it introduced SHAP value to analyze data characteristics. The experimental results show that the fraud account detection model based on XGBoost achieves 0. 205, 0. 084 and 0. 833 in RMSE, MAE and R², which is better than the other comparison models. Combined with the SHAP value, it identifies the key factors that predict fraudulent accounts to provide decision-making reference for the transaction security of blockchain. [ABSTRACT FROM AUTHOR]

The fashion industry is one of the most commercially profitable industries in the world, fashion outfit recommendation technology based on machine learning adopts massive outfit data of the Internet fashion community to learn the outfit matching skills, and realizes personalized outfit recommendation to serve the fashion industry, which has received extensive attentions from both industry and academia. Combining the characteristics of fashion outfit, this paper proposed the fashion outfit recommendation architecture based on machine learning, and on this basis, it summarized and analyzed the relevant research work of key technologies such as compatibility prediction and diagnosis, intelligent outfit generation, and so on. At last, it pointed out the challenging problems and future research directions in this field. [ABSTRACT FROM AUTHOR]

In order to solve the problem that the existing super-resolution algorithm based on deep learning didn't make full use of the feature information of each level, resulting in low reconstruction accuracy and large parameter quantity, this paper proposed a double dense connection structure named densely channel attention skip connection network. In the inner structure of the network, it improved the original dense cascade block to generate a channel separable dense cascade block. The outer structure adopted a densely residual connection and attention mechanism to fuse the features extracted by the dense block to achieve the goal that less convolution layer and higher precision effect. This paper tested the network models on several benchmark datasets . The results show the proposed model has higher accuracy and fewer parameters than the other models. [ABSTRACT FROM AUTHOR]

At present, multi-agent reinforcement learning algorithms mostly adopt frameworks that are centralized in learning and decentralized in action. These frameworks may take too long to converge or may not converge at all. In order to speed up the collective learning time of multi-agents, this paper proposed a novel multi-agent group learning strategy. It used recurrent neural network (RNN) to predict the grouping matrix of multi-agents to share the experience between them, resulting in improved learning efficiency within the multi-agents group. Meanwhile, this paper proposed the concept of information trace to remedy the problem that the agents could not share information brought by the grouping. In order to strengthen the retention of excellent experience within the group, this paper proposed the practice of delaying the death time of excellent agents in the group. Finally, the results show that, compared to MADDPG, the training time is reduced by 12% in the labyrinth experiment and by 17 % in capture the flag experiment. [ABSTRACT FROM AUTHOR]

In order to improve the performance of frame disambiguation model, this paper used a neural network model to learn frame representation based on sentences in corpus different from the traditional classification algorithm extracting features manually, and employed the learned frame representation on the frame disambiguation task, which significantly improved the performance of the task. Making full use of the CFN example sentence database and being based on the hinge-loss neural network, the algorithm learnt the frame representation that could distinguish the correct frame from the error frame in the largest degree. This paper also used the WSABIE algorithm to learn the representation vector of the context of the target word, and finally used the cosine distance between the context representation vector and the frame representation vector to make a decision for the task. Experiment performed three sets of two-fold cross-validation (3 x 2 BCV) on 88 ambiguous words in CFN, and the best accuracy of frame disambiguation reach to 72. 52%. The t-test results show that the performance of the proposed method is significantly higher than other frame disambiguation methods. [ABSTRACT FROM AUTHOR]

Mobile traffic classification/ clustering is an important foundation for mobile network traffic management. However, the mobile network traffic data used by different papers were collected from different network environment. In addition, the labels and the flow statistical features of mobile traffic were different from papers. These experimental results couldn't be directly compared. This paper collected the traffic data generated by App based on MobileGT system. The two kinds of labels were built on these data (App level and function level),and two kinds of flow statistical features were independently extracted on these traffic data. This paper comprehensively researched the machine learning techniques on the traffic data with different labels and different flow statistical features. The experimental results show that the uni-direction flow based features are better than bidirection flow based features, random forest and AdaBoost are better on classifying mobile traffic data, and K-means is better on clustering mobile traffic data. [ABSTRACT FROM AUTHOR]

Based on video action recognition, this paper firstly reviewed the traditional action recognition algorithms based on RGB, including traditional algorithms and deep learning algorithms. Because of the influence of background lighting, RGB-videobased action recognition is low accuracy, but it has rich color appearance information. Then this paper analyzed and summarized action recognition algorithms based on RGB-D, which was divided into depth sequence, bone and multi-feature fusion. RGB-D video has multiple modes and can provide more information for action recognition, which performs better than RGB video but also brings new challenges. Finally, this paper prospected the common data sets and the possible future development direction. [ABSTRACT FROM AUTHOR]

With the advancement of data acquisition technology, spatiotemporal data with geographic location information are growing rapidly, so it is urgent to explore effective spatiotemporal data modeling methods. Spatiotemporal sequence prediction is one of the basic methods in spatiotemporal data modeling and widely used in many fields. At present, there is no Chinese literature on its review, so it is of great significance to summarize these methods. For the spatiotemporal sequence prediction problem, this paper reviewed its application background and development history. Then it introduced some related definitions and characteristics. According to its category, the spatiotemporal sequence prediction methods, based on traditional methods, classical machine learning methods and deep learning methods are analyzed respectively as well as their application scope, advantages and disadvantages. Finally, this paper made a review and prospect over the research directions of spatiotemporal sequence prediction, laying a theoretical foundation for researchers to further study this problem [ABSTRACT FROM AUTHOR]

Abnormal event detection is a hot topic in the field of intelligent surveillance monitoring technology research currently. As an important research content of computer vision, its main goal is to use computers to automatically detect abnormal events. Traditional methods have limitations in the weakness of low-level video feature description ability, high computational cost of anomaly detection methods and poor robustness in modeling complex scenes. In recent years, how to design a high-level semantic feature extraction method, accelerate the process of abnormal event detection, and model complex scenes such as multiple cameras have become the forefront topic of current research. Based on the research situation at home and abroad and the mainstream methods, this paper introduces the basic techniques involved in abnormal event detection in surveillance videos, and analyzes the advantages and disadvantages of various types of surveillance video feature extraction methods， feature learning models and anomaly detection methods. This paper also summarizes the commonly used benchmark datasets that can be used to abnormal event detection in surveillance videos. Finally, we discuss the difficulties, challenges and future development trends of abnormal event detection in surveillance videos. [ABSTRACT FROM AUTHOR]

Aiming at the problem that the existing automatic image description generation models based on visual attention and text attention cannot describe the image details and the whole image at the same time, this paper proposed a model for image captioning which included three sub-modules, ie, an image encoder, an evolutionary neural network, and an adaptive merging decoder. The proposed model could effectively integrate both the visual information and text information, and automatically calculated the proportion of these two information at each time step. The experimental results on two public data sets, Flickr30k and COCO2014, show that the proposed EDLMIC model is superior to other baseline models in four indicators such as METEOR, ROUGE-L, CIDEr and SPICE, and has good performance in a variety of different life scenes. [ABSTRACT FROM AUTHOR]

Aiming at the problems that existing human skeleton action recognition methods couldn't explore sufficient human body information and extract sufficient temporal feature, this paper proposed a model based on posture transformation module and posture fusion module (PTF-SGN), which realized the utilization of the key spatio-temporal information in skeleton diagram. Firstly, by preprocessing the skeleton diagram, the model mined the displacement information of limbs and joints, and extracted the features. Then it used the posture transformation module to obtain the posture adjustment factors from the skeleton image data in an unsupervised learning manner, and adaptively adjusted the body posture to enhance the robustness of the model in different environments. Secondly, it proposed a posture fusion module based on the time attention mechanism, which learned the short-term features and the long-term features, and fused the time characteristics of long and short moments to strengthen the characterization ability of time characteristics. Finally, it extracted the global spatio-temporal feature of the skeleton feature to input into the classification network to obtain the action recognition result. The experimental results on the two 3D skeleton datasets of NTU60 RGB+D and NTU120 RGB+D and the two 2D skeleton datasets of Penn-Action and HARPET show that PTF-SGN model can effectively recognize actions of skeleton time series data. [ABSTRACT FROM AUTHOR]

Existing deep learning based single image super-resolution (SISR) models usually improve the fitting ability of the model by increasing the number of network layers, but fail to fully extract and reuse features, leading low quality of reconstructed images. To solve this problem, this paper proposed an image super-resolution model based on feature fusion and attention mechanism. This model used residual in residual (RIR) structure in feature extraction module. The feature extraction module of the network consisted of several residual groups. Each residual group consisted of several residual blocks. This module implemented local feature fusion in each residual group and global feature fusion between each group. In addition, this model introduced coordinate attention module into each residual block and spatial attention module into each residual group. It verifies that the model is able to fully extract features and reuse features. The final experimental results show that the model is superior to the existing models in objective evaluation indexes and subjective visual effect. [ABSTRACT FROM AUTHOR]

Traffic condition prediction is an important part of intelligent transportation system, and traffic flow is the most direct embodiment of traffic condition. Therefore, traffic flow prediction has important application value. On the one hand, the roads in the city have spatial topological properties, on the other hand, the traffic flow changes dynamically with time. Therefore, the key to the prediction of traffic flow is to model the time and space dependence in the data. In view of this characteristic, this paper used neural network model and attention mechanism to explore the temporal and spatial dependence relationship in traffic flow data, and proposed a traffic flow prediction model based on time map attention. In terms of spatial dependence, it used a learning algorithm combining graph convolution network and attention to assign different weights to nodes with different influence degrees, and added node adaptive learning to effectively extract spatial features. In terms of time dependence, it used the temporal convolution network to extract the temporal features, and expanded the sensing domain by expanding convolution, so as to capture the features of longer time series data. A spatial-temporal network layer was composed of graph attention network and time convolution network, which was finally connected to the output layer to output the prediction results. The model used the combination of graph convolution neural network and attention mechanism to extract spatial features, fully considered the spatial relationship between roads, and used temporal convolution network to capture temporal features. After experiments on two real datasets, it is found that it has good performance in the next 15, 30 and 60 minutes, and the results are better than the existing baselines. [ABSTRACT FROM AUTHOR]

Aiming at the problem of low accuracy of TLBO and easy falling into local optima, the paper proposed a new TLBO algorithm integrating cognitive psychology theory. Incorporating the theory of foot in the door effect in the teaching stage, it could set the periodic learning goals for students with learning difficulties, so as to improve the overall level of students. Adding teacher guidance mechanism in the learning stage was good for increasing the speed of algorithm convergence. There was a self-adjustment stage after learning stage. According to psychological locus of control theory, it divided students into internal control type and external control type. Different types of students adopt different attribution methods for their own performance and took corresponding measures. Using the classic benchmark function to test CPTLBO, the results show that the improved algorithm has advantages in optimization accuracy and convergence speed. Constructing the CPTLBO-ELM tap water supply prediction model and using CPTLBO algorithm to optimize the input weights and hidden layer threshold parameters of the extreme learning machine, so as to improve the prediction accuracy and generalization ability of the model. The simulation results show that it is more accurate for the model which is optimized by CPTLBO algorithm. The simulation results show that it is more accurate for the model which is optimized by CPTLBO algorithm. The simulation results show that it is more accurate for the model which is optimized by CPTLBO algorithm. [ABSTRACT FROM AUTHOR]

This paper proposed an optimization model of deep-knowledge tracking（KTSA-BiGRU） based on self-attention mechanism and bidirectional GRU neural networks owing to the existing deep-knowledge tracking models with weak capture of complex relationships between input exercises and inability to effectively handle long-sequence input data. Firstly, it mapped the learner’s historical learning interaction sequence data to the real value vector sequence. Then, it trained the bidirectional GRU neural network as input to model the learner’s learning process, and finally used, the self-attention mechanism to calculate the probability of the learner correctly answering the next question based on the hidden vectors of the bidirectional GRU neural network output and the attention weight. The performance analysis on the three public datasets can improve the prediction accuracy of deep knowledge tracking. [ABSTRACT FROM AUTHOR]

Aiming at the problem of intra-modality feature representation and inter modality feature fusion in multimodal sentiment analysis, this paper proposed a multi-level hybrid fusion multi-modal sentiment analysis model based on attention mechanism and multi-task learning. Firstly, the model used convolution neural network and bi-directional gated unit to extract the single-modality internal feature. Secondly, it used the cross-modality attention mechanism to realize the pairwise feature fusion between modalities. Thirdly, it used the self-attention mechanism to select the modality Finally, combining with multi-task learning, the model obtained both sentiment and emotion classification results. The experimental results on CMU-MOSEI dataset show that this method can improve the accuracy and F 1-score of sentiment and emotion classification. [ABSTRACT FROM AUTHOR]

In many machine learning applications, the data may be symmetric positive definite (SPD) matrices which are not living in Euclidean space. This paper presented a new probabilistic Riemannian space quantization method based on log-Euclidean metric learning. The proposed method extended the Euclidean probabilistic learning vector quantization (PLVQ) method to deal with SPD matrices by treating them as points on the Riemannian manifold of SPD matrices equipped with log-Euclidean metric, through utilizing a parameterized distance function from log-Euclidean metric learning. On the BCI IV 2 a dataset, the proposed method outperformed Euclidean PLVQ by 20% in terms of recognition accuracy. The proposed method also performs better than the first winner of BCI competition IV on this data set. It obtains comparable classification accuracy to PLVQ using affine invariant Riemannian metric, but requires much less computing time, ieonly needs 1% of the training time, while 10% of the test time. The proposed method also obtains superior performance on the BCI III IIIa and ETH-80 datasets, showing its effectiveness and efficiency. [ABSTRACT FROM AUTHOR]

Real-time bidding( RTE) is a widely used advertising mode in online display advertising. In response to the problem that the best bidding strategy is difficult to obtain due to the high dynamics of the RTE auction environment, this paper proposed a method of optimizing bid strategy based on reinforcement learning (RL), which used the policy optimization with penalized point probability distance (POP3D) algorithm to learn the best bidding strategy. In the POP3D-based bidding framework, the advertising bidding process was modeled as an episodic Markov decision process. Each episode contained a fixed number of time steps, and the bid for each ad display was determined by its estimated click-through rate and bidding factors were jointly determined. At each time step, the bidding agent would adjust the bidding factors according to the auction situation at the previous time step, so that the bidding strategy could adapt to the highly dynamic auction environment. The goal of the bidding agent was to learn the best bidding factor adjustment strategy. The experimental results on the iPin You data set demonstrate that compared with the DRLB algorithm, the proposed bidding algorithm increases the number of clicks by 0. 2 % when the budget ratio is 1/16 and 1/32, as well as increased the win rate by 1. 8%, 1. 0% and 1. 7 % when the budget ratio is were 1/8, 1/16 and 1/32. In addition, the proposed method also has advantages in terms of stability. It shows the superiority of this method. [ABSTRACT FROM AUTHOR]

Joint extraction of entities and relations from unstructured text is an important task in information extraction. The existing methods have achieved considerable performance, but are still subject to some inherent limitations, such as error propagation, redundancy of relation prediction, inability to solve the problem of relations overlap, etc. For this reason, this paper proposed a joint entity relationship extraction model BSGB( BiLSTM + SDA-GAT + BiGCN) based on graph neural network. BSGB was a two-stage predicting process. The first stage of this model extended the semantic dependency analysis to the semantic dependency graph, and proposed a graph attention network to integrate the semantic dependency graph(SDA-GAT). By stacking BiLSTM and SDA-GAT, it extracted sentence sequence and local dependent features, and performed entity span detection and preliminary relationship prediction. In the second stage, it constructed the relation-weighted GCN, which further modeled the interaction between entities and relations, and completed the final extraction of entity-relationship triples. The experimental results on the NYT dataset show that the F1 value of this model reaches 67 .1 %, which is 5. 2% higher than the baseline model in this dataset, and the prediction of the overlap relation is also significantly improved. [ABSTRACT FROM AUTHOR]

Aiming at the order dependency problem of the K2 algorithm, this paper proposed a heuristic algorithm (H-vnK2) to learn the node order effectively from a given data set. Specifically, based on the v-structure knowledge learned by the PC algorithm, it corrected the order of some parent-child nodes in the form of node blocks quickly and accurately, and then obtained the optimal order of some nodes. Based on the neighbor set knowledge learned by the PC algorithm, with the distance threshold heuristic strategy, it further modified the order of the parent-child nodes from the global optimal point of view, and then obtained the best order of all nodes. Experiments show that the proposed algorithm is significantly better than the comparison algorithm on the Asia and Alarm standard data set. Among them, compared with the best-performing method based on causal effects, the proposed algorithm improves the accuracy rate by 7% on average, and the increment can reach up to 33 .3 %. It can learn a more accurate network structure. [ABSTRACT FROM AUTHOR]

Traditional aspect-based sentiment analysis methods lack the research on the interaction between aspect object and its context, which leads to the low accuracy of sentiment analysis. In order to effectively extract richer text features, this paper pro-posed an aspect-based sentiment analysis model (IIMAN) to improve the effect of sentiment polarity judgment, which used multi-head attention mechanism to learn the interaction between aspect object and its context. Firstly, the model used the BERT pre-training model to complete the word vectorization of the input sentences. Then, it used the intra multi-head attention network and the inter multi-head attention network in the attention network to learn the interaction between aspect object and its context, as well as the internal interactions within the context. Finally, it realized the sentiment polarity classification through the pointwise convolution transfer layer, the aspect object attention layer and the output softmax layer. Through the experiments on three public aspect-level sentiment analysis datasets: Twitter, laptop and restaurant, the results show that, compared with other baseline models, IIMAN model achieves better effect in accuracy and F1 value, improves the results of the sentiment polarity classification. [ABSTRACT FROM AUTHOR]

Occlusion in a real environment is one of the main obstacles to accurately analyze and recognize facial expressions. In recent years, researchers have used deep learning technology to solve the problem of high misrecognition rate of facial ex-pressions under occlusion conditions. It mainly summarized the deep learning algorithm of occlusion facial expression recognition and occlusion-related issues. Firstly, this paper summarized the development status of facial expression recognition under partial occlusion conditions, expression representation, and the data set used to study occlusion expression. Secondly, it re-viewed the occlusion expression identify the latest developments in deep learning methods and analyzed the impact of occlusion on expressions. Finally, it summarized the main technical challenges, research difficulties and possible coping strategies. The purpose is to provide a more useful reference and benchmark for future research on occlusion expression recognition. [ABSTRACT FROM AUTHOR]

Iterative learning control problem is studied for a class of discrete nonlinear partial difference systems where the trial lengths vary randomly in this paper. The considered systems are composed of parabolic partial difference equations. Firstly, for the trial lengths vary randomly in systems, a stochastic variable satisfying Bernoulli distribution is introduced. Then, according to the properties of partial difference systems and P-type distributed update schemes, iterative learning controller is designed. Through rigorous theoretical analysis, the convergence sufficient condition of output error under λ norm and contractive mapping principle is established to ensure the convergence of tracking error in the sense of norm. The convergence speed of the algorithm is faster than the classical iterative learning algorithm. Finally, the effectiveness of the algorithm is verified by the numerical simulation. [ABSTRACT FROM AUTHOR]

When the deep reinforcement learning algorithm is trained in a complex dynamic environment, it is difficult for the agent to obtain useful information due to the partial observability of the environment, which leads to typical problems such as failure to learn good strategies and slow algorithm convergence speed. This paper proposed an improved DDPG algorithm based on LSTM and asymmetric actor-critic network. This method introduced the LSTM structure into actor-critic network to learn the hidden states in partially observable Markov through memory reasoning. At the same time, when the actor network only used RGB images as partially observable inputs, the critic network used the complete state of the simulation environment to train, which formed an asymmetric network and speeded up the training convergence. The simulation experiment of manipulator grasping in ROS shows that the proposed algorithm has higher success rate and faster convergence speed compared with DDPG, PPO and LSTM-DDPG. [ABSTRACT FROM AUTHOR]

Because convolutional neural networks( CNN) mostly focus on global feature learning and ignore local feature information containing more details, it is difficult to improve the accuracy of indoor scene recognition. To solve this problem, this paper proposed an indoor scene recognition method based on improved global-local attention network ( GLANet). Firstly, it used GIAN et to capture the global and local features of the scene image to increase the detail information of the image features. Then, it introduced a non-local attention module into the local network to further preserve the detail features of the image through the convolution of attention diagram and feature graph. Finally, it used various features of different stages of the network for classification. Through training and verification on the MIT lndoor67 and the SUN397 dataset, the recognition accuracy of the proposed method was increased by l. 98% and 3. 07% respectively compared with the LGN method. Experimental results show that the algorithm can effectively capture global semantic information and fine spatial details, and significantly improves the recognition accuracy. [ABSTRACT FROM AUTHOR]

With the popularity and development of deep learning, the existence of adversarial samples is a serious threat to the security of the deep learning model. For the attack problem of adversarial samples in the physical world, this paper proposed a physical adversarial sample detection method based on attention mechanism. The method combined attention mechanism and feature compression to detect pertinently of local visual adversarial samples, excluded effects of non-major regions, and reduced calculation effort. It effectively combined multiple feature compression methods to deal with the main areas of the sample, which destroyed the structure of the antagonistic block and made them unaggressive. Perform defense tests on different adversarial attacked on the MNIST and CIFAR-10 data sets and compared with other countermeasures. The experimental results show that the defense accuracy of the method can reach more than 95%. This method is higher versatility and stronger stability than other local adversarial sample defense methods, and can effectively defense the attack of local visual adversarial samples. [ABSTRACT FROM AUTHOR]

The rectangle layout optimization problem with balance constraints is derived from the layout design of satellite module equipment. It belongs to combinatorial optimization problem. Deep reinforcement learning uses reward mechanism to realize high-performance decision optimization through data training. This paper proposed a new deep reinforcement learning algorithm DAR and its extension IDAR. The DAR algorithm output the optimized location sequence with a pointer network, and then used the positioning mechanism to give the layout results. The training of pointer network was realized by deep reinforcement learning. The time complexity of the DAR algorithm was O ( n ³ ). The IDAR algorithm was an iteration version of DAR, which had better results but with a time complexity of 0 ( n 4 ). The test results show that DAR algorithm has good learning ability, and the model obtained by small layout problems can be effectively applied to large-scale problems. The results on two typical large-scaled instances show that the proposed algorithms have achieved or approached to the current best results, so that they have advantages both in time and solution quality. [ABSTRACT FROM AUTHOR]

Implicit aspect extraction is important for improving the accuracy of fine-grained sentiment analysis. However, existing implicit aspect extraction techniques do not have strong generalization ability when dealing with large-scale data. To address the problem, this paper proposed an implicit aspect extraction model combining dependency syntactic parsing and interactive attention mechanism. First, the model generated the initial representation of the text by the pre-trained language model BERT. Then, it passed the initial representation to the self-attention layer guided by the dependency syntactic parsing. Due to the interactive attention mechanism, the model further extracted the results of the above two processes. Finally it used a classifier to determine the implicit aspect of the sentence. Compared with baseline BERT and other deep neural network models, the proposed model has achieved higher F1 and AUC on the enhanced SemEval implicit aspect dataset, which proves the effectiveness of the model. [ABSTRACT FROM AUTHOR]

Memory network only focusing on attention mechanism can't tackle this situation where the context sentiment relies on the concrete target. In order to address this problem, this paper proposed a target-sensitive deep memory network model. This model exploited attention mechanism to capture the sentiment information of context corresponding to a given target. Then the interaction module integrated context sentiment expression and interaction information between context and target to the classification features. Finally, it classified the features to get sentiment polarity of target. This paper conducted experiments on the two datasets from SemEval 2014 task4. The proposed model achieved better performance than attention-based memory networks. The experimental results show that considering the interaction between context and target is effective to solve the problem that context sentiment relies on the concrete target. [ABSTRACT FROM AUTHOR]

In order to enhance the network representation learning quality with the edge signed semantic information, focusing on the weakness of existing fusion methods in dealing with complex edge signed semantic information, this paper proposed a network representation learning algorithm incorporating with edge signed semantic information, and introduced the edge signed semantic information containing positive and negative relations into the network representation learning process. Firstly, this paper designed a relationship prediction model based on the three-layer perceptron to depict different types of context link relations between nodes. Then it introduced the random walk strategy to implement context link sampling to adapt the large-scale network scenarios. Experiments on three data sets show that this algorithm can effectively model different types of context links between nodes and mine the complex semantic information contained in them. Compared with the current optimal SIDE method, the performance of the proposed algorithm is improved by 0. 31 %, 1. 3 % and 1. 85 % . [ABSTRACT FROM AUTHOR]

In sound event detection tasks, the target event was susceptible to background noise, and wasn't present in a significantly high portion of sound signal flow. To solve the problem, this paper proposed a new method of sound event detection based on one-dimensional dense convolutional network (DenseNet ) with multi-level attention mechanism. Firstly, it used the one-dimensional DenseNet for frame-level detection, which was effective in finding the precise onset and offset time. Then, it introduced the multi-level attention mechanism in the one-dimensional DenseNet model, which made the attention-aware features from different modules change adaptively as layers went deeper. Therefore, the model could automatically select and attend on important frames for the targets while ignoring the unrelated parts (e. g. the background noise segments) . Finally, this paper evaluated the model using DCASE 2017 Task 2 development dataset. Results show that the overall performance of the proposed method has further improvement than the conventional deep learning method. [ABSTRACT FROM AUTHOR]

This paper proposed an uncalibrated 3 D robotic arm grabbing method based on improved YOL0v2 in a multi-object environment. Firstly, in order to reduce the depth learning algorithm YOL0v2 detection multi-obj ect bounding box overlapping rate and 3D distance calculation error, it proposed an improved algorithm for YOL0v2. It used this algorithm to detect and identify the target object in the image, obtained the pos ition information of the target object in the RGB image, and then used the K-means + + clustering algorithm to quickly calculate the distance from the target object to the camera according to the depth image information, and estimate the target object size and pose. Simultaneously, it used the improved YOL0v2 to get the bounding box of the gripper and calculated the distance from the robot to the target object. Then the system estimated the distance between the fixture, camera and obj ect in the manipulator coordinate system. Finally, the system used the PID algorithm to control the gripper to grab the object according to the size and posture of the object and the distance from the object to the gripper. The detected boundary boxes of the target obj ect was more accurate based on the improved YOL0v2 than on old one. It also enhanced the distance from the fixture to the object and the size of the object as well as the accuracy of the pose estimation. In addition, in order to avoid complicated calibration, this paper proposed a non-calibration method. The learning scheme was different from the traditional uncalibrated estimation method based on Jacobian matrix, because it had good universality. A simulation experiment shows that the proposed method can accurately classify and locate the objects in the image. The Universal Robot 3 robotic arm uses this framework to verify the effectiveness of capturing objects in a cluttered environment. [ABSTRACT FROM AUTHOR]

This paper proposed constrained deep reinforcement learning( CDRL) to compute the fastest route online using taxi drivers' experience in different time period. Firstly, this paper described the extraction of experiential road segment database (ERSD). Then it introduced CDRL method, which mainly comprised of two phase: bounded condition of route and deep Q-learning algorithm. In the first phase, the task was to generate alternative constrained road segments of OD pair. In the second phase, it devised deep Q-learning algorithm to learning the experience of taxi drivers, and computed the fastest route of a given OD according to their departure time. Lastly, this paper tested an empirical studies in CBD of Guangzhou. The results show that the routes computed by CDRL method is approximately equal to shortest route ( SR) and fastest route (FR) method in travel time and route length. Furthermore, the CDRL method notably outperforms FR and SR in computing efficiency, so it is more suitable for online fastest route computation. [ABSTRACT FROM AUTHOR]

The data sources of existing link prediction algorithms are mainly based on neighbors, paths, and random walk methods. The link prediction algorithms use mainly node similarity assumptions or maximum likelihood estimates. The link prediction based on neural network is still absent. Some research achievements based on neural network show that the Deep Walk algorithm based on neural network is an efficient network representation learning algorithm, which can more effectively learn the network structure features in the network. It has been proven that DeepWalk is equivalent to factorize the target matrix. Therefore, this paper proposed a link prediction algorithm (LPMF) based on matrix factorization of Deep Walk. This algorithm based on matrix factorization used the Deep Walk algorithm to get the network representation vectors. And then, it calculated the similarities between node pairs of nod es by the cosine similarity method. Based on that, the similarity matrix of the target network was constructed. Finally, this paper used the similarity matrix to conduct the link prediction experiments on three real -world citation networks. The experimental results show that the new method is superior to the existing 20 kinds of link prediction algorithms, which fully shows that LPMF can effectively find the structural correlation between nodes in the network, and performs a more excellent performance in the actual tasks of link prediction. [ABSTRACT FROM AUTHOR]

In order to understand the development of SISR based on deep learning and grasp the hotspots and directions of the current research, this paper combed the existing model of SISR based on deep learning. Firstly, the paper introduced the related deep learning algorithm, these models based on deep learning and their evaluation index. In addition, it compared the performance of existing models through experiments, which aimed to understand the advantages of SISR model based on deep learning. Finally, the paper summarized the key issues of SISR, and prospected the future development trends. [ABSTRACT FROM AUTHOR]

This paper proposed a detection algorithm for sweeping jamming with low complexity under the constraint of the probability of false alarm and missed detection, and theoretically analyzed the performance of the detection algorithm to provide the basis for jamming detection of the actual communication system. Then, the paper proposed an anti-jamming learning algorithm based on Q-learning, which could autonomously select the best communication channel and the maximal reliable transmission time when threatened by the sweeping jamming. Finally, the simulation results show that the proposed detection algorithm can effectively detect the linear sweeping jamming, and the detection performance is basically consistent with the theoretical analysis results with low complexity. The anti-jamming learning algorithm can adaptively determine the optimal communication channel, and effectively avoid the sweeping jamming to realize the reliable information transmission. [ABSTRACT FROM AUTHOR]

In speech recognition, the convolution location-based hybrid attention mechanism can not extract location information that can be valid over long term. This paper proposed a new hybrid attention mechanism to solve this problem. Firstly it convolved with the attention score generated for the current time to extract multi-channel features, followed by obtaining the feature vectors of constant dimensions via global average pooling. Then it introduced a LSTM unit as the external memory module and used the generated feature vectors as input to generate the location vectors for the next time point. Finally this paper used the classic LAS model to verify the effect of the new hybrid attention mechanism. Experiment results show that the new hybrid attention mechanism can take full consideration of the attention scores at many past time points. Compared to the convolution location-based LAS model, the label error rate of the proposed scheme on pure and noisy speech datasets is reduced by 1.8% and 2.21%, respectively. [ABSTRACT FROM AUTHOR]

Mixed sampling data，whose different features are collected at different sampling frequencies， pervasively exists in the real world. Because the different sampling data set not only has different features，but also has different number of samples，traditional classification methods cannot be used directly. Therefore， this paper proposes a classification method for mixed sampling data based on Fisher discrimination dictionary learning to solve the classification problem of mixed sampling data. Inspired on some classification ideas for multi-view data，the proposed model compares multiple data collected at multiple sampling frequencies to multiple views of multi-view data，and designs a way to learn a sub-dictionary for each class of each view. A structured dictionary whose dictionary atoms have correspondence to the class labels is learned，so that the within-class scatter is less and the between-class scatter is bigger based on the Fisher discrimination criterion. In addition，this paper designs a specific classification scheme for the inconsistent sample size of mixed sampling data. Finally，experimental results demonstrate the effectiveness of the proposed model. [ABSTRACT FROM AUTHOR]

To solve the problem that lung nodule automatic detection methods for CT images can only get low sensitivities with a lot of false positives, this paper proposed an integrated method with hybrid loss based 3D fully convolutional network as candidate detection and attention-based multi-scale residual network as nodule classification. Firstly, this paper established a 3D fully convolutional network based on dice coefficient loss, and the network filtered hard negative samples uniting with positives to finetune. Then, it designed an attention based multi-scale 3D residual convolutional network to classify the candidate and recognize true nodules. Experiment results on the LUNA16 dataset show that the proposed method achieves the sensitivity of 97. 18% at 59. 1 false positives per scan in the candidate detection stage, and the whole system achieves the average sensitivity of 0. 880, which demonstrates this proposed method can improve sensitivity with fewer false positives and achieve superior performance. [ABSTRACT FROM AUTHOR]

Mental fatigue is a key cause of many chronic diseases such as CVD,diabetes and cancer. It is illusive and hard to measure or detect. This research proposed a feasible accurate and cost saving method to detect fatigue level of mental workers via smart wearable devices. In order to detect mental fatigue level, this paper firstly extracted HRV features, then used Man-Whitney U test to evaluate the statistical significances of HRV features between normal and fatigue state. This paper used random forest for feature selection to determine the importance of HRV features. The most important HRV features were NN. mean, PNN50, VLF, LF and TP respectively. This paper then took four machine learning algorithms on selected features to predict the state of mental fatigue. The experiments demonstrate the effectiveness of KNN classifier with 75. 5% accuracy rate on cross validation and 0.74 AUC. [ABSTRACT FROM AUTHOR]

Ocean mesoscale eddy is an important ocean mesoscale phenomenon that plays an important role in ocean circulation, mat erial and energy transport, and has an important impact on the safety of ship navigation and hydroacoustic communication. Efficient and accurate detection and identification of ocean mesoscale eddies are of great research value for both physical ocean cognition and ocean exploitation. Traditional eddy detection and identification methods rely on a single threshold designed by experts experiences. With the rise of deep learning, the current machine learning methods show certain advantages in the accuracy and automation of eddy detection and identification. This paper summarizes and comparatively analyzes the existing machine learning-based detection and identification methods to provide a systematic knowledge and reference basis for the development of ocean mesoscale eddy detection and identification research. [ABSTRACT FROM AUTHOR]

Objectives: With the rise of location service applications and the popularization of mobile smart terminals, the demand for high ‑ precision navigation and positioning is extending from isolated areas to seamless whole areas, from professional groups to mass users. However, real‑time, continuous, and complete pedestrian navigation still faces many challenges, such as occlusion of global navigation satellite system (GNSS) signals, frequent updates of Wi‑Fi/ bluetooth low energy (BLE)/geomagnetic fingerprint data‑ bases, and smoothly switching in the transitional area of indoor and outdoor. In addition, the positioning sensors of popular mobile smart terminals are limited by cost and power consumption, and the observation data is generally noisy and poor in stability. Methods: Therefore, this paper proposes a pedestrian indoor and outdoor seamless positioning technology based on cloud‑end collaboration of smartphone, which breaks through the crowd sourced collection and rapid update of the beacon fingerprint database, the precise positioning of the smart phone multi‑sensor information fusion, and the indoor and outdoor seamless switching. We have developed a smartphone positioning application for pedestrians. The collaborative precision positioning enhancement information is provided by the location service platform and the multi‑source observations data (GNSS, miniature inertial measurement unit (MIMU), Wi ‑ Fi, BLE, magnetometer, barometer) are obtained by the smartphone. Results: The feild test results show that the outdoor positioning accuracy of ordinary users is better than 1.5 m, and indoor accuracy is about 1 ‒ 3 m. Conclusions: It supports the demand of mass‑market users for seamless and precise positioning indoors and outdoors. [ABSTRACT FROM AUTHOR]

Cloud segmentation is a key task in applications of Landsat8 remote sensing images. At present, most existing deep learning based approaches usually adopt the U-type coder-encoder network structure. Although this structure can make use of spatial location information from the coding end effectively, the whole network has too many parameters and cause high computation cost. Meanwhile, the coding end only adopts simple convolution and subsampling operation, which makes it impossible to obtain high-quality contextual semantic feature information effectively. In view of the above situation, this paper proposed a new lightweight U-Net model for cloud segmentation. The whole model adopted a new U-shaped codec structure that jumped and connected the information between the shallow layer and the middle layer at the coding end, and also embedded novel optimization blocks consist of grouping convolution and attention mechanism in coding end. It also built context-semantic fusion connection to link corresponding layers in coding and decoding end. Experimental results on the common benchmark dataset 38-Cloud illustrates, compared with other mainstream cloud segmentation networks, the proposed new U-Net can achieve better segmentation precision with less model parameters. [ABSTRACT FROM AUTHOR]

Aiming at the problems of complex background, variety of change types, missing detection and rough boundary recognition in high-resolution remote sensing image of the same region, this paper proposed a high-resolution remote sensing image building change detection algorithm based on Siam-UNet++network. The algorithm used UNet++as the backbone extraction network. In the encoder phase, it applied the Siam-diff structure to extract the change features of the two sequential images, and employed the attention mechanism after the up sampling and lateral jump path connection in the decoding stage to highlight the building change features and inhibit the network learning from other types of features. Meanwhile, it used the MSOF strategy to weight and fuse feature information of different semantic levels, which improved the accuracy of building change detection. Finally, it adopted a sliding window method to predict large-scale remote sensing images, reducing the hole pattern generated by the change result map during the splicing process. The experimental results demonstrate that proposed algorithm shows better performance than other models. [ABSTRACT FROM AUTHOR]

It is critical to obtain sufficient and effective EEG features in mental workload evaluation. This paper presented a multiclass classification network which combined multi-branch LSTM with attention mechanism. Firstly, this network segmented an EEG signal into a number of temporal slices, in each of which multi-branch LSTM extracted temporal feature from the EEG slices. Then attention mechanism optimized the features. Finally, softmax function evaluated mental workload level. The experimental results show that the performance of the proposed network is better than that of the existing network in both binary class and multiclass classification task. [ABSTRACT FROM AUTHOR]