Search

Stochastic point location deals with the problem of finding a target point on a real line through a learning mechanism (LM) with the stochastic environment (SE) offering directional information. The SE can be further categorized into an informative or deceptive one, according to whether p is above 0.5 or not, where p is the probability of providing a correct suggestion of a direction to LM. Several attempts have been made for LM to work in both types of environments, but none of them considers a dynamically changing environment where p varies with time. A dynamic dual environment involves fierce changes that frequently cause its environment to switch from an informative one to a deceptive one, or vice versa. This article presents a novel weak estimator-based adaptive step search solution, to enable LM to track the target in a dynamic dual environment, with the help of a weak estimator. The experimental results show that the proposed solution is feasible and efficient.

Representation learning with small labeled data have emerged in many problems, since the success of deep neural networks often relies on the availability of a huge amount of labeled data that is expensive to collect. To address it, many efforts have been made on training sophisticated models with few labeled data in an unsupervised and semi-supervised fashion. In this paper, we will review the recent progresses on these two major categories of methods. A wide spectrum of models will be categorized in a big picture, where we will show how they interplay with each other to motivate explorations of new ideas. We will review the principles of learning the transformation equivariant, disentangled, self-supervised and semi-supervised representations, all of which underpin the foundation of recent progresses. Many implementations of unsupervised and semi-supervised generative models have been developed on the basis of these criteria, greatly expanding the territory of existing autoencoders, generative adversarial nets (GANs) and other deep networks by exploring the distribution of unlabeled data for more powerful representations. We will discuss emerging topics by revealing the intrinsic connections between unsupervised and semi-supervised learning, and propose in future directions to bridge the algorithmic and theoretical gap between transformation equivariance for unsupervised learning and supervised invariance for supervised learning, and unify unsupervised pretraining and supervised finetuning. We will also provide a broader outlook of future directions to unify transformation and instance equivariances for representation learning, connect unsupervised and semi-supervised augmentations, and explore the role of the self-supervised regularization for many learning problems., Comment: published in IEEE Transactions on Pattern Analysis and Machine Intelligence

Differentiable architecture search (DARTS) enables effective neural architecture search (NAS) using gradient descent, but suffers from high memory and computational costs. In this paper, we propose a novel approach, namely Partially-Connected DARTS (PC-DARTS), to achieve efficient and stable neural architecture search by reducing the channel and spatial redundancies of the super-network. In the channel level, partial channel connection is presented to randomly sample a small subset of channels for operation selection to accelerate the search process and suppress the over-fitting of the super-network. Side operation is introduced for bypassing (non-sampled) channels to guarantee the performance of searched architectures under extremely low sampling rates. In the spatial level, input features are down-sampled to eliminate spatial redundancy and enhance the efficiency of the mixed computation for operation selection. Furthermore, edge normalization is developed to maintain the consistency of edge selection based on channel sampling with the architectural parameters for edges. Theoretical analysis shows that partial channel connection and parameterized side operation are equivalent to regularizing the super-network on the weights and architectural parameters during bilevel optimization. Experimental results demonstrate that the proposed approach achieves higher search speed and training stability than DARTS. PC-DARTS obtains a top-1 error rate of 2.55 percent on CIFAR-10 with 0.07 GPU-days for architecture search, and a state-of-the-art top-1 error rate of 24.1 percent on ImageNet (under the mobile setting) within 2.8 GPU-days.

In this paper, the H ∞ sliding mode control (SMC) problem is discussed for a class of uncertain discrete networked systems with channel fading and randomly varying nonlinearities (RVNs) by employing the observer-based approach. Here, the Jth-order Rice fading model is adopted to describe the phenomenon of channel fading, where the channel coefficients are mutually independent random variables that take the values on certain interval. In addition, a sequence of mutually independent Bernoulli distributed random variables is utilized to characterize the phenomenon of the RVNs, where the occurrence probabilities of RVNs consist of the nominal mean and the bound of error. The main objective of the addressed problem is to design the observer-based sliding mode controller such that the closed-loop system is asymptotically stable in mean-square sense and the pre-specified H ∞ performance requirement is guaranteed. By employing novel Lyapunov–Krasovskii functional, sufficient criteria are derived to force the system state trajectories onto a pre-specified sliding mode region. Afterwards, the solvability of non-convex problem is discussed by testifying the feasibility of a minimization problem. Finally, an illustrative example is given to demonstrate the effectiveness of aforementioned SMC technique.

This paper proposes a novel tensor-train deep neural network (TT-DNN) based channel estimator to tackle challenges of time-varying channel estimation in multiple-input multiple-output (MIMO) systems. A centralized DNN channel estimator can be realized by a distributed TT-DNN with parallel paths. The TT-DNN provides a compact representation by decomposing each DNN layer into a TT format with fewer model parameters, and is well-designed to adapt to the block structure, pilot density, and the number of MIMO antennas. Moreover, the channel estimation is performed in block-by-block and antenna-by-antenna manners such that both input dimensions of TT-DNN and the number of model parameters can be further reduced. In addition, a preliminary stage of model pre-training is set for the DNN/TT-DNN channel estimator which boosts the channel estimation accuracy. Moreover, the proposed TT-DNN is generalized to semiblind channel estimation scenarios where there exists only preamble training symbols. Our experiments show that the proposed TT-DNN based channel estimator outperforms the DNN counterparts in terms of convergence rate, estimation accuracy, and robustness, and presents better performance than recurrent neural network for semiblind channel estimation.

In this work, we aim to address the problem of human interaction recognition in videos by exploring the long-term inter-related dynamics among multiple persons. Recently, Long Short-Term Memory (LSTM) has become a popular choice to model individual dynamic for single-person action recognition due to its ability to capture the temporal motion information in a range. However, most existing LSTM-based methods focus only on capturing the dynamics of human interaction by simply combining all dynamics of individuals or modeling them as a whole. Such methods neglect the inter-related dynamics of how human interactions change over time. To this end, we propose a novel Hierarchical Long Short-Term Concurrent Memory (H-LSTCM) to model the long-term inter-related dynamics among a group of persons for recognizing human interactions. Specifically, we first feed each person's static features into a Single-Person LSTM to model the single-person dynamic. Subsequently, at one time step, the outputs of all Single-Person LSTM units are fed into a novel Concurrent LSTM (Co-LSTM) unit, which mainly consists of multiple sub-memory units, a new cell gate, and a new co-memory cell. In the Co-LSTM unit, each sub-memory unit stores individual motion information, while this Co-LSTM unit selectively integrates and stores inter-related motion information between multiple interacting persons from multiple sub-memory units via the cell gate and co-memory cell, respectively. Extensive experiments on several public datasets validate the effectiveness of the proposed H-LSTCM by comparing against baseline and state-of-the-art methods.

Human parsing, especially in the wild, has attracted a lot of attention due to its great potential in many real-world applications. The Pyramid Spatial Parsing (PSP) module has shown superior performances in scene and human parsing tasks. However, the basic AvgPool operation in PSP equally aggregates spatial clues of a local region, and thus mixes up influences of different human parts presented in this region. It results in failures in capturing useful contexts relevant to parsing different parts. To address this problem, a suitable mechanism to collect spatial clues aligning with different human parts is proposed in this paper. We employ a Gather-Excite (GE) operation, a replacement of the AvgPool-Upsample operation in a pyramidical structure, to accurately reflect relevant human parts of various scales. The GE operation contains two steps: the gather operation that adaptively aggregates spatial clues to relevant human parts, and the excite operation that generates new feature maps with the gathered contextual information. This results in a novel Pyramidical Gather-Excite Context (PGEC) module to solve the multi-scale problem and parse person at various scales. The PGEC module is composed of multiple GE operations with different spatial extents and aggregates local and global spatial clues for better modeling multi-scale contextual information in parallel. Moreover, we integrate the PGEC module with fine-grained details, edge preserving module and deep supervision to formulate a novel PGEC Network (PGECNet) for human parsing. The proposed PGECNet has achieved state-of-the-art performance on four single-person human parsing datasets ( i.e. , LIP, PPSS, ATR and Fashion Clothing) and two multi-person human parsing datasets ( i.e. , PASCAL-Person-Part and CIHP). The experimental results show that the proposed PGEC is superior to the PSP and ASPP modules especially in single-human parsing task. The source code is publicly available at https://github.com/31sy/PGECNet .

Prone to problems flow calculation does not converge at the scheduled maximum power operating mode, the current operating personnel inability to visually observe the system state, convergence can only be adjusted by trial and error approach. Solving the problem of non convergence of power flow plays an important role in power system analysis. Based on the intermediate process of solving the Newton method, the concept of an approximate fashion, the relation between convergence and voltage stability is calculated by analyzing the trend of the voltage characteristics similar trend as the main basis of the judgment result in the trend does not converge, based on voltage sensitivity further, by improving the convergence of the method to boost the voltage level indirectly. Finally, after the New England 39-bus system to verify, by demonstrating the relevance and accuracy of the proposed method. When not solve the problem of the convergence trend is not impressive, it has a certain reference value for practical application. This study has a good application prospect in practical engineering application

Representation learning methods allow a system to automatically learn robust and discriminative features from raw data for given goals, which play an important role in various visual content understanding applications, such as visual object segmentation, detection, tracking, recognition, and search. The performance of visual content understanding tasks is heavily dependent on the choice of data representation (or features) on which they are applied. Conventional feature representation methods usually employ transformations of data that make it easier to extract useful information, such as scale-invariant feature transform (SIFT), local binary patterns (LBP), and histogram of oriented gradients (HOG). In recent years, deep learning techniques have been widely applied to learn data-driven representations with supervised annotations and achieved great success in different visual content understanding tasks. Representative methods include the ResNet method for image classification, the DeepFace method for face recognition, and the feature pyramid networks (FPNs) method for object detection. Despite recent progresses on deep representation learning with a great amount of annotated data, how to effectively learn visual representation with limited data annotations still requires many efforts. This special section focuses on data-effective representation learning methods for visual content understanding.

Recent emergence of “microblogging” services has been driving cyber-physical social system (CPSS) as a hot topic in real-world applications. How to efficiently detect and recognise spam and fake accounts becomes an important task where it requires analysis of microblog user behavior and prediction of their activity. This article attempts to investigate this challenge by proposing a new strategy to effectively model microblogging user activity and dynamically predicting their activities for the CPSS applications. We first analysis and define a set of benchmarks for measuring microblogging user activeness in considering serval key dynamic attributes including change rate of microblogging numbers, user attentions, etc. Then, we build up a new dynamic microblogging user activity prediction model (DUAPM) based on three important characteristics: personal information, social relationship, and user interaction. Finally, an improved logical regression algorithm is proposed for training the model and predicting user activity. Under the evaluation of a sample dataset containing Sina Weibo 3621 users over 20 weeks, it shows that our model deliver average up to 3% higher prediction accuracy than other social media user activity prediction models using traditional logical regression and random forest algorithms. We also take out a CPSS case study of evaluating DUAPM models for analysis and prediction of Twitter users’ activity over 16 countries. The results show that our model effectively reflects the distribution and trends of Twitter users’ activity with different background and cultures.

In visual object tracking, by reasonably fusing multiple experts, ensemble framework typically achieves superior performance compared to the individual experts. However, the necessity of parallelly running all the experts in most existing ensemble frameworks heavily limits their efficiency. In this paper, we propose POST, a POlicy-based Switch Tracker for robust and efficient visual tracking. The proposed POST tracker consists of multiple weak but complementary experts (trackers) and adaptively assigns one suitable expert for tracking in each frame. By formulating this expert switch in consecutive frames as a decision-making problem, we learn an agent via reinforcement learning to directly decide which expert to handle the current frame without running others. In this way, the proposed POST tracker maintains the performance merit of multiple diverse models while favorably ensuring the tracking efficiency. Extensive ablation studies and experimental comparisons against state-of-the-art trackers on 5 prevalent benchmarks verify the effectiveness of the proposed method.

Collaborative estimation of a sparse vector $\mathbf {x}$ x by $M$ M potential measurements is considered. Each measurement is the projection of $\mathbf {x}$ x obtained by a regressor, i.e., $y_m=\mathbf {a}_m^T \mathbf {x}$ y m = a m T x . The problem of selecting $K$ K sensor measurements from a set of $M$ M potential sensors is studied where $K\ll M$ K ≪ M and $K$ K is less than the dimension of $\mathbf {x}$ x . In other words, we aim to reduce the problem to an under-determined system of equations in which a sparse solution is desired. This paper suggests selecting sensors in a way that the reduced matrix construct a well conditioned measurement matrix. Our criterion is based on E-optimality, which is highly related to the restricted isometry property that provides some guarantees for sparse solution obtained by $\ell _1$ l 1 minimization. The proposed basic E-optimal selection is vulnerable to outlier and noisy data. The robust version of the algorithm is presented for distributed selection for big data sets. Moreover, an online implementation is proposed that involves partially observed measurements in a sequential manner. Our simulation results show the proposed method outperforms the other criteria for collaborative spectrum sensing in cognitive radio networks (CRNs). Our suggested selection method is evaluated in machine learning applications. It is used to pick up the most informative features/data. Specifically, the proposed method is exploited for face recognition with partial training data.

Due to importantly beneficial effects on physical and mental health and strong association with many rehabilitation programs, Physical Activity Recognition and Measure (PARM) has been widely recognised as a key paradigm for a variety of smart healthcare applications. Traditional methods for PARM relies on designing and utilising Data fusion or machine learning techniques in processing ambient and wearable sensing data for classifying types of physical activity and removing their uncertainties. Yet they mostly focus on controlled environments with the aim of increasing types of identifiable activity subjects, improved recognition accuracy and measure robustness. The emergence of the Internet of Things (IoT) enabling technology is transferring PARM studies to an open and dynamic uncontrolled ecosystem by connecting heterogeneous cost-effective wearable devices and mobile apps and various groups of users [35] . Little is currently known about whether traditional Data fusion techniques can tackle new challenges of IoT environments and how to effectively harness and improve these technologies. In an effort to understand potential use and opportunities of Data fusion techniques in IoT enabled PARM applications, this paper will give a systematic review, critically examining PARM studies from a perspective of a novel 3D dynamic IoT based physical activity collection and validation model. It summarized traditional state-of-the-art data fusion techniques from three plane domains in the 3D dynamic IoT model: devices, persons and timeline. The paper goes on to identify some new research trends and challenges of data fusion techniques in the IoT enabled PARM studies, and discusses some key enabling techniques for tackling them.

The state-of-the-art machine learning approaches are based on classical von Neumann computing architectures and have been widely used in many industrial and academic domains. With the recent development of quantum computing, researchers and tech-giants have attempted new quantum circuits for machine learning tasks. However, the existing quantum computing platforms are hard to simulate classical deep learning models or problems because of the intractability of deep quantum circuits. Thus, it is necessary to design feasible quantum algorithms for quantum machine learning for noisy intermediate scale quantum (NISQ) devices. This work explores variational quantum circuits for deep reinforcement learning. Specifically, we reshape classical deep reinforcement learning algorithms like experience replay and target network into a representation of variational quantum circuits. Moreover, we use a quantum information encoding scheme to reduce the number of model parameters compared to classical neural networks. To the best of our knowledge, this work is the first proof-of-principle demonstration of variational quantum circuits to approximate the deep $Q$-value function for decision-making and policy-selection reinforcement learning with experience replay and target network. Besides, our variational quantum circuits can be deployed in many near-term NISQ machines., Accepted for publication by IEEE Access

A wide-area damping controller (WADC) is effective in damping inter-area low-frequency oscillation (LFO), if the time delay in a wide-area control loop can be properly handled. In order to simplify the WADC design and enlarge the delay adaptation range, the classic power system stabilizer (PSS) is adopted, and a new unified residue (UR) method is proposed for compact WADC design. The strategy of control loop selection is also improved by modifying the relative residue index based on a few dominant oscillation modes. The designed PSS-based compact WADC is as simple as classic PSS with no more than two lead-lag phase compensation units. Case studies are carried out on an IEEE 16-machine 68-bus power system. Simulation results demonstrate that the control loop selection before the WADC design is necessary and that the proposed selection strategy can easily pick out the suitable candidate control loops. In addition, it is feasible for the UR method to design WADCs with different time delays in the selected control loops. All the designed WADCs are effective in damping inter-area LFO and robust to time delay variations under operation conditions. Comparisons among five design methods for PSS-based WADC show that the proposed UR method is superior in delay adaptation, the conciseness of WADC structure and computation speed of parameters.

Domestic waste classification was incorporated into legal provisions recently in China. However, relying on manpower to detect and classify domestic waste is highly inefficient. To that end, in this article, we propose a multimodel cascaded convolutional neural network (MCCNN) for domestic waste image detection and classification. MCCNN combined three subnetworks (DSSD, YOLOv4, and Faster-RCNN) to obtain the detections. Moreover, to suppress the false-positive predicts, we utilized a classification model cascaded with the detection part to judge whether the detection results are correct. To train and evaluate MCCNN, we designed a large-scale waste image dataset (LSWID), containing 30 000 domestic waste multilabeled images with 52 categories. To the best of our knowledge, the LSWID is the largest dataset on domestic waste images. Furthermore, a smart trash can is designed and applied to a Shanghai community, which helped to make waste recycling more efficient. Experimental results showed a state-of-the-art performance, with an average improvement of 10% in detection precision.

BACKGROUND: Multiparametric dual-energy computed tomography (mpDECT) is widely used to differentiate various kinds of tumors; however, the data regarding its diagnostic performance with machine learning to diagnose breast tumors is limited. We evaluated univariate analysis and machine learning performance with mpDECT to distinguish between benign and malignant breast lesions. METHODS: In total, 172 patients with 214 breast lesions (55 benign and 159 malignant) who underwent preoperative dual-phase contrast-enhanced DECT were included in this retrospective study. Twelve quantitative features were extracted for each lesion, including CT attenuation (precontrast, arterial, and venous phases), the arterial-venous phase difference in normalized effective atomic number (nZ(eff)), normalized iodine concentration (NIC), and slope of the spectral Hounsfield unit (HU) curve (λ(Hu)). Predictive models were developed using univariate analysis and eight machine learning methods [logistic regression, extreme gradient boosting (XGBoost), stochastic gradient descent (SGD), linear discriminant analysis (LDA), adaptive boosting (AdaBoost), random forest (RF), decision tree, and linear support vector machine (SVM)]. Classification performances were assessed based on the area under the receiver operating characteristic curve (AUROC). The best performances of the conventional univariate analysis and machine learning methods were compared using the Delong test. RESULTS: The univariate analysis showed that the venous phase λ(Hu) had the highest AUROC (0.88). Machine learning with mpDECT achieved an excellent and stable diagnostic performance, as shown by the mean classification performances in the training dataset (AUROC, 0.88–0.99) and testing (AUROC, 0.83–0.96) datasets. The performance of the AdaBoost model based on mpDECT was more stable than the other machine learning models and superior to the univariate analysis (AUROC, 0.96 vs. 0.88; P

This paper focuses on a theoretical analysis of deep neural network (DNN) based functional approximation. Leveraging upon two classical theorems on universal approximation, an artificial neural network (ANN) with a single hidden layer of neurons is used. With modified ReLU and Sigmoid activation functions, we first generalize the related concepts to vector-to-vector regression. Then, we show that the width of the hidden layer of ANN is numerically related to the approximation of the regression function. Furthermore, we increase the number of hidden layers and show that the depth of the ANN-based regression function can enhance its expressive power. We illustrate this representation with recently-emerged DNN based speech enhancement. We first compare the expressive power by varying ANN structures and then test its related regression performance under different noisy conditions in various noise types and signal-to-noise-ratio levels. Experimental results verify our theoretical prediction that an ANN of a broader hidden layer and a deeper architecture can jointly ensure a closer approximation of the vector-to-vector regression functions in terms of the Euclidean distance between the log power spectra of noisy and expected clean speech. Moreover, a DNN with a broader width at the top hidden layer can improve the regression performance relative to those with a narrower width at the top hidden layers.

As a new revolution of Ubiquitous Computing and Internet of Things, multimodal wearable intelligence technique is rapidly becoming a new research topic in both academic and industrial fields. Owning to the rapid spread of wearable and mobile devices, this technique is evolving healthcare from traditional hub-based systems to more personalised healthcare systems. This trend is well-aligned with recent “Healthcare 4.0” which is a continuous process of transforming the entire healthcare value chain to be preventive, precise, predictive and personalised, with significant benefits to elder care. But empowering the utility of multimodal wearable intelligence technique for elderly care like people with dementia is significantly challenging considering many issues, such as shortage of cost-effective wearable sensors, heterogeneity of wearable devices connected, high demand for interoperability, etc. Focusing on these challenges, this paper gives a systematic review of advanced multimodal wearable intelligence technologies for dementia care in Healthcare 4.0. One framework is proposed for reviewing the current research of wearable intelligence, and key enabling technologies, major applications, and successful case studies in dementia care, and finally points out future research trends and challenges in Healthcare 4.0.

Before the implementation of the task, reviewing the status of key facilities and equipment is an important activity to reduce the risk of failure and ensure the success of the task. If the number of equipment participating in the task is large and the review time is limited, the review personnel need to carry out the review work with focus and purpose according to the actual situation of the task. According to the actual needs of reexamination work of reexamination staff, this paper puts forward the comprehensive evaluation index system of equipment quality before task, defines the specific meaning and calculation method of each index, constructs the index weight calculation model based on analytic hierarchy process, calculates the index weight, and gives an application example. This work provides a reference for equipment quality evaluation ranking before task and reexamination personnel to determine reexamination focus.

We present a novel attention mechanism: Causal Attention (CATT), to remove the ever-elusive confounding effect in existing attention-based vision-language models. This effect causes harmful bias that misleads the attention module to focus on the spurious correlations in training data, damaging the model generalization. As the confounder is unobserved in general, we use the front-door adjustment to realize the causal intervention, which does not require any knowledge on the confounder. Specifically, CATT is implemented as a combination of 1) In-Sample Attention (IS-ATT) and 2) Cross-Sample Attention (CS-ATT), where the latter forcibly brings other samples into every IS-ATT, mimicking the causal intervention. CATT abides by the Q-K-V convention and hence can replace any attention module such as top-down attention and self-attention in Transformers. CATT improves various popular attention-based vision-language models by considerable margins. In particular, we show that CATT has great potential in large-scale pre-training, e.g., it can promote the lighter LXMERT~\cite{tan2019lxmert}, which uses fewer data and less computational power, comparable to the heavier UNITER~\cite{chen2020uniter}. Code is published in \url{https://github.com/yangxuntu/catt}.

As the rapid growth of social media technologies continues, Cyber-Physical-Social System (CPSS) has been a hot topic in many industrial applications. The use of “microblogging” services, such as Twitter, has rapidly become an influential way to share information. While recent studies have revealed that understanding and modelling microblog user behaviour with massive users’ data in social media are keen to success of many practical applications in CPSS, a key challenge in literatures is that diversity of geography and cultures in social media technologies strongly affect user behaviour and activity. The motivation of this article is to understand differences and similarities between microblogging users from different countries using social media technologies, and to attempt to design a Country-Level Micro-Blog User (CLMB) behaviour and activity model for supporting CPSS applications. We proposed a CLMB model for analysing microblogging user behaviour and their activity across different countries in the CPSS applications. The model has considered three important characteristics of user behaviour in microblogging data, including content of microblogging messages, user emotion index, and user relationship network. We evaluated CLBM model under the collected microblog dataset from 16 countries with the largest number of representative and active users in the world. Experimental results show that (1) for some countries with small population and strong cohesiveness, users pay more attention to social functionalities of microblogging service; (2) for some countries containing mostly large loose social groups, users use microblogging services as a news dissemination platform; (3) users in countries whose social network structure exhibits reciprocity rather than hierarchy will use more linguistic elements to express happiness in microblogging services.

Recent advancements in recurrent neural network (RNN) research have demonstrated the superiority of utilizing multiscale structures in learning temporal representations of time series. Currently, most of multiscale RNNs use fixed scales, which do not comply with the nature of dynamical temporal patterns among sequences. In this paper, we propose Adaptively Scaled Recurrent Neural Networks (ASRNN), a simple but efficient way to handle this problem. Instead of using predefined scales, ASRNNs are able to learn and adjust scales based on different temporal contexts, making them more flexible in modeling multiscale patterns. Compared with other multiscale RNNs, ASRNNs are bestowed upon dynamical scaling capabilities with much simpler structures, and are easy to be integrated with various RNN cells. The experiments on multiple sequence modeling tasks indicate ASRNNs can efficiently adapt scales based on different sequence contexts and yield better performances than baselines without dynamical scaling abilities.

In this article, we propose a novel deep Siamese architecture based on a convolutional neural network (CNN) and multi-level similarity perception for the person re-identification (re-ID) problem. According to the distinct characteristics of diverse feature maps, we effectively apply different similarity constraints to both low-level and high-level feature maps during training stage. Due to the introduction of appropriate similarity comparison mechanisms at different levels, the proposed approach can adaptively learn discriminative local and global feature representations, respectively, while the former is more sensitive in localizing part-level prominent patterns relevant to re-identifying people across cameras. Meanwhile, a novel strong activation pooling strategy is utilized on the last convolutional layer for abstract local-feature aggregation to pursue more representative feature representations. Based on this, we propose final feature embedding by simultaneously encoding original global features and discriminative local features. In addition, our framework has two other benefits: First, classification constraints can be easily incorporated into the framework, forming a unified multi-task network with similarity constraints. Second, as similarity-comparable information has been encoded in the network’s learning parameters via back-propagation, pairwise input is not necessary at test time. That means we can extract features of each gallery image and build an index in an off-line manner, which is essential for large-scale real-world applications. Experimental results on multiple challenging benchmarks demonstrate that our method achieves splendid performance compared with the current state-of-the-art approaches.

Many researches have brought progress in learning a good generative model by combing the advantages of GAN and VAE, where latent space learning is always important for generating high-quality images. But these existing works mainly seek to impose the latent space a given distribution in advance or to obey a Gaussian distribution with KL divergence penalty, which leads to the difficulties of deciding a suitable prior distribution corresponding to different datasets. Thus in this paper we develop a two-stage combining method of AE and GAN under unsupervised and supervised conditions respectively, each stage performed to improve the effect of modeling latent distribution. In the first stage, an adversarial procedure achieves to match the latent distribution with real data distribution determined by arbitrary dataset without having access to a pre-set prior distribution. In the second stage, besides one adversarial procedure trained for outputting images, the other adversarial procedure is designed to attain the goal of optimizing the latent distribution of the first stage via back-propagation. Therefore, loop optimization of the network parameters during training will at last allow the framework to map an input noise to a high quality image. Extensive experiments are conducted to verify the performance of representing latent space and generating images on different datasets including MNIST, fashion-MNIST, CIFAR-10, and CelebA. The code and tutorials already release at https://github.com/TwistedW/CAE-CGAN .

Hashing has attracted increasing research attention in recent years due to its high efficiency of computation and storage in image retrieval. Recent works have demonstrated the superiority of simultaneous feature representations and hash functions learning with deep neural networks. However, most existing deep hashing methods directly learn the hash functions by encoding the global semantic information, while ignoring the local spatial information of images. The loss of local spatial structure makes the performance bottleneck of hash functions, therefore limiting its application for accurate similarity retrieval. In this paper, we propose a novel deep ordinal hashing (DOH) method, which learns ordinal representations to generate ranking-based hash codes by leveraging the ranking structure of feature space from both local and global views. In particular, to effectively build the ranking structure, we propose to learn the rank correlation space by exploiting the local spatial information from fully convolutional network and the global semantic information from the convolutional neural network simultaneously. More specifically, an effective spatial attention model is designed to capture the local spatial information by selectively learning well-specified locations closely related to target objects. In such hashing framework, the local spatial and global semantic nature of images is captured in an end-to-end ranking-to-hashing manner. Experimental results conducted on three widely used datasets demonstrate that the proposed DOH method significantly outperforms the state-of-the-art hashing methods.

Cross-resolution face recognition (CRFR), which is important in intelligent surveillance and biometric forensics, refers to the problem of matching a low-resolution (LR) probe face image against high-resolution (HR) gallery face images. Existing shallow learning-based and deep learning-based methods focus on mapping the HR-LR face pairs into a joint feature space where the resolution discrepancy is mitigated. However, little works consider how to extract and utilize the intermediate discriminative features from the noisy LR query faces to further mitigate the resolution discrepancy due to the resolution limitations. In this study, we desire to fully exploit the multi-level deep convolutional neural network (CNN) feature set for robust CRFR. In particular, our contributions are threefold. (i) To learn more robust and discriminative features, we desire to adaptively fuse the contextual features from different layers. (ii) To fully exploit these contextual features, we design a feature set-based representation learning (FSRL) scheme to collaboratively represent the hierarchical features for more accurate recognition. Moreover, FSRL utilizes the primitive form of feature maps to keep the latent structural information, especially in noisy cases. (iii) To further promote the recognition performance, we desire to fuse the hierarchical recognition outputs from different stages. Meanwhile, the discriminability from different scales can also be fully integrated. By exploiting these advantages, the efficiency of the proposed method can be delivered. Experimental results on several face datasets have verified the superiority of the presented algorithm to the other competitive CRFR approaches., IEEE Transactions on Circuits and Systems for Video Technology, 11 pages, 9 figures

Human motion prediction aims to generate future motions based on the observed human motions. Witnessing the success of Recurrent Neural Networks (RNN) in modeling the sequential data, recent works utilize RNN to model human-skeleton motion on the observed motion sequence and predict future human motions. However, these methods did not consider the existence of the spatial coherence among joints and the temporal evolution among skeletons, which reflects the crucial characteristics of human motion in spatiotemporal space. To this end, we propose a novel Skeleton-joint Co-attention Recurrent Neural Networks (SC-RNN) to capture the spatial coherence among joints, and the temporal evolution among skeletons simultaneously on a skeleton-joint co-attention feature map in spatiotemporal space. First, a skeleton-joint feature map is constructed as the representation of the observed motion sequence. Second, we design a new Skeleton-joint Co-Attention (SCA) mechanism to dynamically learn a skeleton-joint co-attention feature map of this skeleton-joint feature map, which can refine the useful observed motion information to predict one future motion. Third, a variant of GRU embedded with SCA collaboratively models the human-skeleton motion and human-joint motion in spatiotemporal space by regarding the skeleton-joint co-attention feature map as the motion context. Experimental results on human motion prediction demonstrate the proposed method outperforms the related methods.

When traditional inventory methods sort out the data, the data format is not standardized, which leads to the long query time of inventory data. To solve this problem, this paper proposes an enterprise level cross domain integrated database storage method based on DMAIC model. Through the collection of cross domain integrated data for filtering and integration, the data are transferred to the central database in turn, and the data warehouse is established. The core data in the collected data is identified by using DMAIC model, and the core data is summarized into the inventory range. Then, the relevant data with high similarity are matched, and labels and data catalog are made to provide support for data positioning and query. The results show that compared with the traditional method, the query time of inventory data is shortened by 0.57s.

In the detection of genome variation, the research on the internal correlation of reference genome is deepening; the detection of variation in genome sequence has become the focus of research, and it has also become an effective path to find new genes and new functional proteins. The targeted sequencing sequence is used to sequence the exon region of a specific gene in cancer gene detection, and the sequencing depth is relatively large. Traditional alignment algorithms will lose some sequences, which will lead to inaccurate mutation detection. This paper proposes a mutation detection algorithm based on feedback fast learning neural network position index. By establishing a position index relationship for ACGT in the DNA sequence, the subsequence is decomposed into the position relationship of different subsequences corresponding to the main sequence. The positional relationship of the subsequence in the main sequence is determined by the positional relationship. Analyzing SNP and InDel mutations, even structural mutations, through the position correlation of sequences has the advantages of high precision and easy implementation by personal computers. The feedback fast learning neural network is used to verify whether there is a linear relationship between two or more positions. Experimental results show that the mutation points detected by position index are more than those detected by Bcftools, Freebye, Vanscan2, and Gatk.

Aiming at the problem that the current visualization analysis method of power enterprise data assets does not consider the design of data visualization analysis model, which leads to the low accuracy of the line chart formed by visualization, this paper proposes a data assets hot spot visualization analysis method based on data platform. Starting from the three problems of data assets, such as lack, error and repetition, this paper preprocesses the data assets of power enterprises, adopts the technology of data middle platform, establishes the framework of data middle platform to mine the hot spots of data assets of power enterprises. This paper designed a data visualization analysis model, build the visualization system structure of electric power enterprise data assets hot, implements the hot visualization of electric power enterprise data assets. The experimental results show that the line chart formed by the visualization analysis method of power enterprise data assets hot spot has high accuracy.

The importance of the Physical Activity (PA) study in the uncertain, dynamic environment of daily life has been drawn much attention. Traditional researches demonstrate that leveraging advanced on-body wearable sensors and effective machine learning algorithms enable monitoring and recognizing human daily PA in controlled environments with high accuracy. But they usually suffer from low accuracy and weak robustness of objectively qualifying PA in daily life due to restrictions like less-attachments, low sensitivity of consumable wearables and subject's variation. Thus, there is a lack of holistic investigation on how to improve accuracy of PA recognition in daily life using cost-effective wearable devices with feasible algorithms. In this paper, we use less-attached on-body consumable wearables like belt and wristband devices to collect data from two different groups of people(10 healthy people and 7 mild cognitive impairment patients). We compared the features from different types of sensors (accelerometer, gyroscope and magnetometer) and performed feature fusion. Then, by using Artificial Neural Network(ANN) to classify the features after fusion, we get a cost-effective wearable intelligence approaches for PA recognition in daily life. Parameters of achieving high recognition rate including time window sizes, features and activation functions. The experimental results indicate that PA containing 10 subjects collected by wrist in daily life environments with average accuracy up to 93%. In addition, ANN had good results in 7 mild cognitive impairment (MCI) patients, reaching 96%.

Driven by the new revolution in healthcare, the importance of understanding Physical Activity (PA) in the uncertain, dynamic, free-living environments has been drawing growing attention. However, there is a lack of holistic investigation on how to improve the accuracy of PA recognition in free-living environments using cost-effective wearable devices with feasible algorithms. In this paper, we design a two-layer hybrid hierarchical model for accessing and evaluating cost-effective wearable intelligence approaches for PA recognition in free-living environments. The hypothesis of this model first suggests utilising less-attached on-body consumable wearables like belt and wristband devices, and then building up a PA dataset collected in free-living environments like an elderly home, hospital, office and gym. The model is then defined with components of lightGBM (LGB) and Artificial Neural Networks (ANN) for coarse and fine-grained classification, parameters of achieving high recognition rate including time window sizes, features and activation functions. The experimental results indicate that our model has superior ability over other state-of-the-art algorithms in classifying three typical types of PA (dynamic, sedentary, and transitional) with an average accuracy up to 84%. Specifically, our model performs good results of PA recognition with the ageing populations including 5 Mild Cognitive Impairment (MCI) and 17 Parkinson's disease (PD) patients.

The prediction and modeling of chronic diseases such as Alzheimer's disease (AD) has received widespread attention in recent years. The field is tightly integrated with medical care, and recent advances in machine learning technology provide opportunities to train AD disease progression models. This trend has led to the exploration and design of new machine learning techniques for multimodal medical and health datasets to predict the occurrence and modeling process of AD. The purpose of this article is to perform a longitudinal and tracking analysis by machine learning models and explore core regions of the brain's components associated with AD progression that are important for brain degeneration. We summarized the biomarkers related with the progression of AD and checked them by neuropathology. On the basis of this, further generalization to the corresponding functional brain blocks was provided; the results showed that this is in line with the results in the field of clinical medicine and provides technical support for physician-assisted diagnosis.

Wind power grid-connected capacity continues to grow rapidly in our country, the continuous and safe grid connection of wind turbines after the grid is affected by disturbances is more and more important to the stable operation of the system. Paper conducts research based on the above status, firstly, the reactive power control strategy model of GE-type doubly-fed induction generator (DFIG) that packaged in PSD-BPA is analyzed, and the DFIG grid-connected test system is established. By using the large disturbance excitation method, the influence of changes in grid-connected bus voltage on the transient response of DFIG is studied. Aiming at the increase process after a sharp drop in voltage, paper reveals that the internal potential of wind turbines with voltage feedback control function has the characteristic of “callback lag”, analyzes the mechanism of this characteristic triggering the threat of DFIG high-voltage trip-off, and puts forward the countermeasures. Finally, combined with the simulation results of short-circuit fault in Hami area power grid, paper verifies the existence of the threat of high-voltage trip-off under such failures and the effectiveness of the proposed countermeasures to reduce the threat.

Transformation Equivariant Representations (TERs) aim to capture the intrinsic visual structures that equivary to various transformations by expanding the notion of translation equivariance underlying the success of Convolutional Neural Networks (CNNs). For this purpose, we present both deterministic AutoEncoding Transformations (AET) and probabilistic AutoEncoding Variational Transformations (AVT) models to learn visual representations from generic groups of transformations. While the AET is trained by directly decoding the transformations from the learned representations, the AVT is trained by maximizing the joint mutual information between the learned representation and transformations. This results in Generalized TERs (GTERs) equivariant against transformations in a more general fashion by capturing complex patterns of visual structures beyond the conventional linear equivariance under a transformation group. The presented approach can be extended to (semi-)supervised models by jointly maximizing the mutual information of the learned representation with both labels and transformations. Experiments demonstrate the proposed models outperform the state-of-the-art models in both unsupervised and (semi-)supervised tasks. Moreover, we show that the unsupervised representation can even surpass the fully supervised representation pretrained on ImageNet when they are fine-tuned for the object detection task.

Background: With the advent of the second generation sequencing technology, the discovery of sequence alignment and sequence variation is a long-standing challenge. Results: A method based on general alignment software, SNP and Indel software tools was proposed in this paper. By comparing the advantages of traditional alignment software, we can produce the best alignment. SNP and Indel detection tools fusion research found that different depth of sequencing effect is different. When the sequence depth reaches a certain value, select one of the software for testing. Conclusions: Finally, the experimental verification shows that SNP and Indel have certain advantages in the comparison of the effects of the fusion method.

In this paper, we present a novel regularization mechanism for training deep networks by minimizing the {\em Worse-Case Perturbation} (WCP). It is based on the idea that a robust model is least likely to be affected by small perturbations, such that its output decisions should be as stable as possible on both labeled and unlabeled examples. We will consider two forms of WCP regularizations -- additive and DropConnect perturbations, which impose additive noises on network weights, and make structural changes by dropping the network connections, respectively. We will show that the worse cases of both perturbations can be derived by solving respective optimization problems with spectral methods. The WCP can be minimized on both labeled and unlabeled data so that networks can be trained in a semi-supervised fashion. This leads to a novel paradigm of semi-supervised classifiers by stabilizing the predicted outputs in presence of the worse-case perturbations imposed on the network weights and structures.

Generative Adversarial Networks (GAN) have shown promising performance in image synthesis and unsupervised learning (USL). In most cases, however, the representations extracted from unsupervised GAN are usually unsatisfactory in other computer vision tasks. By using conditional GAN (CGAN), this problem could be solved to some extent, but the main drawback of such models is the necessity for labeled data. To improve both image synthesis quality and representation learning performance under the unsupervised setting, in this paper, we propose a simple yet effective Transformation Generative Adversarial Networks (TrGAN). In our approach, instead of capturing the joint distribution of image-label pairs p(x,y) as in conditional GAN, we try to estimate the joint distribution of transformed image t(x) and transformation t. Specifically, given a randomly sampled transformation t, we train the discriminator to give an estimate of input transformation, while following the adversarial training scheme of the original GAN. In addition, intermediate feature matching as well as feature-transform matching methods are introduced to strengthen the regularization on the generated features. To evaluate the quality of both generated samples and extracted representations, extensive experiments are conducted on four public datasets. The experimental results on the quality of both the synthesized images and the extracted representations demonstrate the effectiveness of our method.

Recent advances in Graph Convolutional Neural Networks (GCNNs) have shown their efficiency for non-Euclidean data on graphs, which often require a large amount of labeled data with high cost. It it thus critical to learn graph feature representations in an unsupervised manner in practice. To this end, we propose a novel unsupervised learning of Graph Transformation Equivariant Representations (GraphTER), aiming to capture intrinsic patterns of graph structure under both global and local transformations. Specifically, we allow to sample different groups of nodes from a graph and then transform them node-wise isotropically or anisotropically. Then, we self-train a representation encoder to capture the graph structures by reconstructing these node-wise transformations from the feature representations of the original and transformed graphs. In experiments, we apply the learned GraphTER to graphs of 3D point cloud data, and results on point cloud segmentation/classification show that GraphTER significantly outperforms state-of-the-art unsupervised approaches and pushes greatly closer towards the upper bound set by the fully supervised counterparts. The code is available at: https://github.com/gyshgx868/graph-ter.

If the sprinkler system cannot be started at the time of need, it may cause huge economic losses. The system needs to have high reliability. Therefore, in the design stage of sprinkler system, it is necessary to carry out reliability modeling and prediction to ensure that the system reliability can meet the requirements. In this paper, the design scheme of a building sprinkler system is analyzed in depth, and then a task reliability model considering maintenance strategy, stand-by period failure and task period failure is established, and the reliability is predicted to evaluate whether the design scheme of the system can meet the design needs.

Building energy consumption data plays an important role in building energy analysis and energy saving optimization. However, due to difficulties in collecting, high cost, equipment failure and other reasons, the collected data are prone to be missing, which hinders the mining and analysis of building energy consumption data. In this paper, the Bayesian network is used to check and recover the building energy consumption data. In the case that the amount of time series data missing is less than 50%, the method G-test is selected to identify abnormal data, and the Naive Bayesian optimizing Expected Maximum Algorithm is used to check the data. When a large number of building energy consumption data missing, the Sparse Bayesian learning algorithm is used to fill in the missing data based on the compressed sensing theory. The results show that the model can effectively deal with the problem of missing data of building energy consumption and can be widely used in practical projects.

We propose a tensor-to-vector regression approach to multi-channel speech enhancement in order to address the issue of input size explosion and hidden-layer size expansion. The key idea is to cast the conventional deep neural network (DNN) based vector-to-vector regression formulation under a tensor-train network (TTN) framework. TTN is a recently emerged solution for compact representation of deep models with fully connected hidden layers. Thus TTN maintains DNN's expressive power yet involves a much smaller amount of trainable parameters. Furthermore, TTN can handle a multi-dimensional tensor input by design, which exactly matches the desired setting in multi-channel speech enhancement. We first provide a theoretical extension from DNN to TTN based regression. Next, we show that TTN can attain speech enhancement quality comparable with that for DNN but with much fewer parameters, e.g., a reduction from 27 million to only 5 million parameters is observed in a single-channel scenario. TTN also improves PESQ over DNN from 2.86 to 2.96 by slightly increasing the number of trainable parameters. Finally, in 8-channel conditions, a PESQ of 3.12 is achieved using 20 million parameters for TTN, whereas a DNN with 68 million parameters can only attain a PESQ of 3.06. Our implementation is available online https://github.com/uwjunqi/Tensor-Train-Neural-Network., Comment: Accepted to ICASSP 2020. Update reproducible code

Recent studies have highlighted adversarial examples as ubiquitous threats to the deep neural network (DNN) based speech recognition systems. In this work, we present a U-Net based attention model, U-Net$_{At}$, to enhance adversarial speech signals. Specifically, we evaluate the model performance by interpretable speech recognition metrics and discuss the model performance by the augmented adversarial training. Our experiments show that our proposed U-Net$_{At}$ improves the perceptual evaluation of speech quality (PESQ) from 1.13 to 2.78, speech transmission index (STI) from 0.65 to 0.75, short-term objective intelligibility (STOI) from 0.83 to 0.96 on the task of speech enhancement with adversarial speech examples. We conduct experiments on the automatic speech recognition (ASR) task with adversarial audio attacks. We find that (i) temporal features learned by the attention network are capable of enhancing the robustness of DNN based ASR models; (ii) the generalization power of DNN based ASR model could be enhanced by applying adversarial training with an additive adversarial data augmentation. The ASR metric on word-error-rates (WERs) shows that there is an absolute 2.22 $\%$ decrease under gradient-based perturbation, and an absolute 2.03 $\%$ decrease, under evolutionary-optimized perturbation, which suggests that our enhancement models with adversarial training can further secure a resilient ASR system., The authors have revised some annotations in Table 4 to improve the clarity. The authors thank reading feedbacks from Jonathan Le Roux. The first draft was finished in August 2019. Accepted to IEEE ICASSP 2020

Facial expression and identity are two independent yet intertwined components for representing a face. For facial expression recognition, identity can contaminate the training procedure by providing tangled but irrelevant information. In this paper, we propose to learn clearly disentangled and discriminative features that are invariant of identities for expression recognition. However, such disentanglement normally requires annotations of both expression and identity on one large dataset, which is often unavailable. Our solution is to extend conditional VAE to a crossed version named Cross-VAE, which is able to use partially labeled data to disentangle expression from identity. We emphasis the following novel characteristics of our Cross-VAE: (1) It is based on an independent assumption that the two latent representations’ distributions are orthogonal. This ensures both encoded representations to be disentangled and expressive. (2) It utilizes a symmetric training procedure where the output of each encoder is fed as the condition of the other. Thus two partially labeled sets can be jointly used. Extensive experiments show that our proposed method is capable of encoding expressive and disentangled features for facial expression. Compared with the baseline methods, our model shows an improvement of 3.56% on average in terms of accuracy.

Recent deep neural networks based techniques, especially those equipped with the ability of self-adaptation in the system level such as deep reinforcement learning (DRL), are shown to possess many advantages of optimizing robot learning systems (e.g., autonomous navigation and continuous robot arm control.) However, the learning-based systems and the associated models may be threatened by the risks of intentionally adaptive (e.g., noisy sensor confusion) and adversarial perturbations from real-world scenarios. In this paper, we introduce timing-based adversarial strategies against a DRL-based navigation system by jamming in physical noise patterns on the selected time frames. To study the vulnerability of learning-based navigation systems, we propose two adversarial agent models: one refers to online learning; another one is based on evolutionary learning. Besides, three open-source robot learning and navigation control environments are employed to study the vulnerability under adversarial timing attacks. Our experimental results show that the adversarial timing attacks can lead to a significant performance drop, and also suggest the necessity of enhancing the robustness of robot learning systems., Accepted to IEEE ICASSP 2020

Distributed automatic speech recognition (ASR) requires to aggregate outputs of distributed deep neural network (DNN)-based models. This work studies the use of submodular functions to design a rank aggregation on score-based permutations, which can be used for distributed ASR systems in both supervised and unsupervised modes. Specifically, we compose an aggregation rank function based on the Lovasz Bregman divergence for setting up linear structured convex and nested structured concave functions. The algorithm is based on stochastic gradient descent (SGD) and can obtain well-trained aggregation models. Our experiments on the distributed ASR system show that the submodular rank aggregation can obtain higher speech recognition accuracy than traditional aggregation methods like Adaboost. Code is available online~\footnote{https://github.com/uwjunqi/Subrank}., Accepted to ICASSP 2020. Please download the pdf to view Figure 1. arXiv admin note: substantial text overlap with arXiv:1707.01166

Physical activity (PA) is widely recognized as one of the important elements of personal healthy life. To date, as the development of wearable sensing technologies, it is possible to utilize wearable devices and machine learning algorithms to efficiently and accurately monitor PA types, intensity and its associated human pattern for many health applications. But there is a trade-off between less-attachment of wearable devices and achievement of high accuracy in existing PA recognition studies. This paper attempts to investigate possible utilisation of Artificial Neural Networks (ANN) achieving high recognition accuracy of PA using less-attachments of wearable devices. Following a four-steps designed experimental protocol, we collect the real activities dataset with only belt and wristband devices from 10 healthy subjects at home and gym environment. The parameters of typical PA recognition with ANN including time window sizes, features and activation functions are evaluated under 24 different subjects of activities. The experimental results indicate that ANN dealing with belt and wristband data can achieve satisfactory PA recognition results in dynamic and sedentary activities but suffers from transitional activities in both environments.

The traditional power carrier wave machine based on digital theory cannot meet the needs of large distance and stable transmission of actual power signals. Therefore, it is necessary to design power carrier wave machine based on TCP/IP model of computer network communication. According to the reference model of TCP/IP for computer network communication, it designs the overall hardware circuit of carrier based on network interface layer, including network layer, transport layer and application layer. The network layer solves the problem of carrier signal attenuation and improves the effective communication distance by exploring the parent node networking algorithm and routing strategy. It uses OFDM technology to demodulate carrier signal and improve the stability of carrier communication. The experimental results show that the designed carrier wave machine achieves data transmission and collection. In addition, the networking algorithm at the network layer can efficiently construct the network, reduce the packet loss rate, improve the communication stability, and enhance the signal modulation ability.

Township is the weak point and weak point in the coordinated development of China’s economy. It is also a link that must be strengthened in the coordinated development of China’s economy. The development of power grid provides a strong guarantee for the development of township economy and related industries. With the smooth progress of “SG – ERP project” of state grid corporation of China, the informatization construction of township power supply units has made remarkable achievements, and the informatization and datamation of daily management have been basically realized. However, there still exists the construction pattern of insufficient intelligence and failure to combine cloud computing technology to support the network of power supply system, which also greatly depresses the ideal implementation effect of information system. In order to fundamentally solve these constraints, it is necessary to carry out a series of power supply team integrated information system construction. In the current power supply system, the team is the direct executor of the daily work of the state grid corporation, as well as the grass-roots organization of the entire power industry. The efficiency of the township power supply station and the team directly affects the reputation of the state grid corporation. This paper will first make a detailed description of the information status of the township power supply station, and then put forward the overall framework of the township power supply station information integration construction system in this study, then the system is deployed and tested, and finally the application value and efficiency of the system are analyzed.