Convolutional Neural Network Module Research Articles

Integrating Convolution and Sparse Coding for Learning Low-Dimensional Discriminative Image Representations.

This work investigates the problem of efficiently learning discriminative low-dimensional (LD) representations of multiclass image objects. We propose a generic end-to-end approach that jointly optimizes sparse dictionary and convolutions for learning LOW-dimensional discriminative image representations, named SparConvLow, taking advantage of convolutional neural networks (CNNs), dictionary learning, and orthogonal projections. The whole learning process can be summarized as follows. First, a CNN module is employed to extract high-dimensional (HD) preliminary convolutional features. Second, to avoid the high computational cost of direct sparse coding on HD CNN features, we learn sparse representation (SR) over a task-driven dictionary in the space with the feature being orthogonally projected. We then exploit the discriminative projection on SR. The whole learning process is consistently treated as an end-to-end joint optimization problem of trace quotient maximization. The cost function is well-defined on the product of the CNN parameters space, the Stiefel manifold, the Oblique manifold, and the Grassmann manifold. By using the explicit gradient delivery, the cost function is optimized via a geometrical stochastic gradient descent (SGD) algorithm along with the chain rule and the backpropagation. The experimental results show that the proposed method can achieve a highly competitive performance with the state-of-the-art (SOTA) image classification, object categorization, and face recognition methods, under both supervised and semi-supervised settings. The code is available at https://github.com/MVPR-Group/SparConvLow.

Enhancing sewage flow prediction using an integrated improved SSA-CNN-Transformer-BiLSTM model

<p>Accurate prediction of sewage flow is crucial for optimizing sewage treatment processes, cutting down energy consumption, and reducing pollution incidents. Current prediction models, including traditional statistical models and machine learning models, have limited performance when handling nonlinear and high-noise data. Although deep learning models excel in time series prediction, they still face challenges such as computational complexity, overfitting, and poor performance in practical applications. Accordingly, this study proposed a combined prediction model based on an improved sparrow search algorithm (SSA), convolutional neural network (CNN), transformer, and bidirectional long short-term memory network (BiLSTM) for sewage flow prediction. Specifically, the CNN part was responsible for extracting local features from the time series, the Transformer part captured global dependencies using the attention mechanism, and the BiLSTM part performed deep temporal processing of the features. The improved SSA algorithm optimized the model's hyperparameters to improve prediction accuracy and generalization capability. The proposed model was validated on a sewage flow dataset from an actual sewage treatment plant. Experimental results showed that the introduced Transformer mechanism significantly enhanced the ability to handle long time series data, and an improved SSA algorithm effectively optimized the hyperparameter selection, improving the model's prediction accuracy and training efficiency. After introducing an improved SSA, CNN, and Transformer modules, the prediction model's $ {R^{\text{2}}} $ increased by 0.18744, $ RMSE $ (root mean square error) decreased by 114.93, and $ MAE $ (mean absolute error) decreased by 86.67. The difference between the predicted peak/trough flow and monitored peak/trough flow was within 3.6% and the predicted peak/trough flow appearance time was within 2.5 minutes away from the monitored peak/trough flow time. By employing a multi-model fusion approach, this study achieved efficient and accurate sewage flow prediction, highlighting the potential and application prospects of the model in the field of sewage treatment.</p>

Enhancing sewage flow prediction using an integrated improved SSA-CNN-Transformer-BiLSTM model

<p>Accurate prediction of sewage flow is crucial for optimizing sewage treatment processes, cutting down energy consumption, and reducing pollution incidents. Current prediction models, including traditional statistical models and machine learning models, have limited performance when handling nonlinear and high-noise data. Although deep learning models excel in time series prediction, they still face challenges such as computational complexity, overfitting, and poor performance in practical applications. Accordingly, this study proposed a combined prediction model based on an improved sparrow search algorithm (SSA), convolutional neural network (CNN), transformer, and bidirectional long short-term memory network (BiLSTM) for sewage flow prediction. Specifically, the CNN part was responsible for extracting local features from the time series, the Transformer part captured global dependencies using the attention mechanism, and the BiLSTM part performed deep temporal processing of the features. The improved SSA algorithm optimized the model's hyperparameters to improve prediction accuracy and generalization capability. The proposed model was validated on a sewage flow dataset from an actual sewage treatment plant. Experimental results showed that the introduced Transformer mechanism significantly enhanced the ability to handle long time series data, and an improved SSA algorithm effectively optimized the hyperparameter selection, improving the model's prediction accuracy and training efficiency. After introducing an improved SSA, CNN, and Transformer modules, the prediction model's $ {R^{\text{2}}} $ increased by 0.18744, $ RMSE $ (root mean square error) decreased by 114.93, and $ MAE $ (mean absolute error) decreased by 86.67. The difference between the predicted peak/trough flow and monitored peak/trough flow was within 3.6% and the predicted peak/trough flow appearance time was within 2.5 minutes away from the monitored peak/trough flow time. By employing a multi-model fusion approach, this study achieved efficient and accurate sewage flow prediction, highlighting the potential and application prospects of the model in the field of sewage treatment.</p>

MSDSE: Predicting drug-side effects based on multi-scale features and deep multi-structure neural network

Unexpected side effects may accompany the research stage and post-marketing of drugs. These accidents lead to drug development failure and even endanger patients' health. Thus, it is essential to recognize the unknown drug-side effects. Most existing methods in silico find the answer from the association network or similarity network of drugs while ignoring the drug-intrinsic attributes. The limitation is that they can only handle drugs in the maturation stage.To be suitable for early drug-side effect screening, we conceive a multi-structural deep learning framework, MSDSE, which synthetically considers the multi-scale features derived from the drug. MSDSE can jointly learn SMILES sequence-based word embedding, substructure-based molecular fingerprint, and chemical structure-based graph embedding. In the preprocessing stage of MSDSE, we project all features to the abstract space with the same dimension. MSDSE builds a bi-level channel strategy, including a convolutional neural network module with an Inception structure and a multi-head Self-Attention module, to learn and integrate multi-modal features from local to global perspectives. Finally, MSDSE regards the prediction of drug-side effects as pair-wise learning and outputs the pair-wise probability of drug-side effects through the inner product operation. MSDSE is evaluated and analyzed on benchmark datasets and performs optimally compared to other baseline models. We also set up the ablation study to explain the rationality of the feature approach and model structure. Moreover, we select model partial prediction results for the case study to reveal actual capability. The original data are available at http://github.com/yuliyi/MSDSE.

A Novel Unsupervised Segmentation Method of Canopy Images from UAV Based on Hybrid Attention Mechanism

Crown segmentation is a pivotal process in the acquisition of tree parameters. In light of the high expenses associated with satellite remote sensing and LiDAR technologies, our study leverages the cost-effective and efficient UAV remote sensing technology for capturing crown images. In addition, considering the expense and sensitivity associated with labeling data for supervised learning and its implications on model generalization and label quality, this paper introduces an innovative unsupervised learning framework based on convolutional neural networks (CNN). To address the limited receptive field of CNN, we have introduced a novel hybrid attention module following each CNN module. This enhancement ensures the integrity of the segmentation results and the coherence of the boundaries. Furthermore, in response to the growing need for user interaction, we have incorporated a scribble interaction function. Through the semantic segmentation of the collected crown images, our proposed method attains remarkable results, achieving an accuracy of 98.15%, an F1_score of 97.01%, and an mIoU of 95.58%. Additionally, we have conducted a comparative analysis of our proposed method with two clustering algorithms, namely K-Means and GMM, and two CNN models, DeepLab and U-Net. The results reveal that our segmentation structures outperform other methods significantly. The experimental findings demonstrate the immense application potential of this method in diverse fields, including forestry management, environmental protection, and ecosystem monitoring.

Hybrid Convolution neural Network and Graph Neural network for Detection of Leukocyte

In pathology and medical diagnostics, the identification of leukocytes, white blood cells essential for immune system function, is significant. The hybrid neural network model used in this study, which combines the strengths of convolutional neural networks (CNNs) and graph neural networks (GNNs), gives a unique method for the identification of leukocytes. Due to the intricate spatial and relational patterns visible in microscopic pictures of blood samples, leukocyte identification is difficult. The proposed hybrid model incorporates a GNN module to record the contextual associations among leukocytes in a blood sample and a CNN module to extract fine-grained characteristics from individual leukocyte pictures. Each leukocyte picture is seen in this paradigm as a node in a network, with the edges denoting the interactions between the cells' locations or contexts. While the GNN module uses graph-based representations to capture the connections and dependencies between leukocytes within a blood sample, the CNN module efficiently extracts local visual characteristics from leukocyte pictures. These two modules are used to produce a thorough and context-sensitive depiction of the leukocyte detection issue. The importance of this study resides in its potential to increase the precision and effectiveness of leukocyte identification in pathology and medical diagnostics, leading to better patient care and illness diagnosis. This hybrid CNN-GNN model may also be used for a variety of other medical image analysis tasks that call for taking both local and contextual information into account. In successfully identifying leukocytes in microscopic images, this hybrid CNN-GNN technique shows promising results. Accuracy, precision, recall, and F1-score are common assessment metrics used to assess the model's performance.

Enhanced Feature Extraction Network Based on Acoustic Signal Feature Learning for Bearing Fault Diagnosis.

The method of acoustic radiation signal detection not only enables contactless measurement but also provides comprehensive state information during equipment operation. This paper proposes an enhanced feature extraction network (EFEN) for fault diagnosis of rolling bearings based on acoustic signal feature learning. The EFEN network comprises four main components: the data preprocessing module, the information feature selection module (IFSM), the channel attention mechanism module (CAMM), and the convolutional neural network module (CNNM). Firstly, the one-dimensional acoustic signal is transformed into a two-dimensional grayscale image. Then, IFSM utilizes three different-sized convolution filters to process input image data and fuse and assign weights to feature information that can attenuate noise while highlighting effective fault information. Next, a channel attention mechanism module is introduced to assign weights to each channel. Finally, the convolutional neural network (CNN) fault diagnosis module is employed for accurate classification of rolling bearing faults. Experimental results demonstrate that the EFEN network achieves high accuracy in fault diagnosis and effectively detects rolling bearing faults based on acoustic signals. The proposed method achieves an accuracy of 98.52%, surpassing other methods in terms of performance. In comparative analysis of antinoise experiments, the average accuracy remains remarkably high at 96.62%, accompanied by a significantly reduced average iteration time of only 0.25 s. Furthermore, comparative analysis confirms that the proposed algorithm exhibits excellent accuracy and resistance against noise.

Joint Measurement of Thermal Discomfort by Occupant Pose, Motion and Appearance in Indoor Surveillance Videos for Building Energy Saving

To accurately describe the thermal comfort of indoor occupant used to regulate heating, ventilation and air conditioning (HVAC) system is a key goal to reduce energy consumption in intelligent buildings. In this paper, we propose a noncontact measurement of occupant thermal discomfort behavior as an index of thermal comfort. The method takes the existing monitoring image data in the building as the input to infer the occupant thermal discomfort directly, which saves the cost because there is no need to install new sensors in the building and the occupant does not need to wear additional equipment. The framework combined three channels of body posture, motion and performance information to infer occupant thermal discomfort behavior, it consists of a human detection and crop module, a posture analysis module, an optical flow extraction module and a 3D convolutional neural network module. The three channels describe the actions from different perspectives, and the proposed method makes full use of the complementarity of the three modalities to identify the occupant’s thermal discomfort behaviors. Sixteen postures related to thermal discomfort were identified through a questionnaire, and 14,800 video clips containing these postures were collected for experimental evaluation. The results demonstrate the superior performances of our approach to the state-of-the-art techniques. The framework achieves noninvasive, cost-effective thermal comfort evaluation, and has potential value in improving energy efficiency of HVAC systems.

SGNet: Sequence-Based Convolution and Ligand Graph Network for Protein Binding Affinity Prediction.

Protein-ligand binding can play an important role in many fields. It is of great importance to accurately predict the binding affinity between molecules by computational methods. Most computational binding affinity methods require molecular structures. However, there are still a large number of protein molecules with known amino acid sequences whose structures have not yet been solved. To address this issue, this paper proposes a sequence-based convolution and ligand graph network, called SGNet, to fuse the molecular graph information and the amino acid sequence information. This method integrates Conjoint Triad (CT) encoding of amino acid sequence and one-dimensional convolutional neural network module to extract protein molecules, develops graph attention network to extract molecular features of ligand, and then fuses the two feature sets to predict the binding affinity between molecules from the fully connected layer. As a result, SGNet achieves good prediction performance on both KIKD and IC50 data sets, with prediction error RMSEs of 1.287 and 1.58, and correlation Pearson Rs of 0.687 and 0.592, respectively. Comparative experimental results under the same conditions showed that SGNet outperformed Kdeep and GraphDTA in predicting binding affinities between protein-ligand molecules.

SeriesSleepNet: an EEG time series model with partial data augmentation for automatic sleep stage scoring.

Introduction: We propose an automatic sleep stage scoring model, referred to as SeriesSleepNet, based on convolutional neural network (CNN) and bidirectional long short-term memory (bi-LSTM) with partial data augmentation. We used single-channel raw electroencephalography signals for automatic sleep stage scoring. Methods: Our framework was focused on time series information, so we applied partial data augmentation to learn the connected time information in small series. In specific, the CNN module learns the time information of one epoch (intra-epoch) whereas the bi-LSTM trains the sequential information between the adjacent epochs (inter-epoch). Note that the input of the bi-LSTM is the augmented CNN output. Moreover, the proposed loss function was used to fine-tune the model by providing additional weights. To validate the proposed framework, we conducted two experiments using the Sleep-EDF and SHHS datasets. Results and Discussion: The results achieved an overall accuracy of 0.87 and 0.84 and overall F1-score of 0.80 and 0.78 and kappa value of 0.81 and 0.78 for five-class classification, respectively. We showed that the SeriesSleepNet was superior to the baselines based on each component in the proposed framework. Our architecture also outperformed the state-of-the-art methods with overall F1-score, accuracy, and kappa value. Our framework could provide information on sleep disorders or quality of sleep to automatically classify sleep stages with high performance.

Rice disease identification method based on improved CNN-BiGRU

In the field of precision agriculture, diagnosing rice diseases from images remains challenging due to high error rates, multiple influencing factors, and unstable conditions. While machine learning and convolutional neural networks have shown promising results in identifying rice diseases, they were limited in their ability to explain the relationships among disease features. In this study, we proposed an improved rice disease classification method that combines a convolutional neural network (CNN) with a bidirectional gated recurrent unit (BiGRU). Specifically, we introduced a residual mechanism into the Inception module, expanded the module's depth, and integrated an improved Convolutional Block Attention Module (CBAM). We trained and tested the improved CNN and BiGRU, concatenated the outputs of the CNN and BiGRU modules, and passed them to the classification layer for recognition. Our experiments demonstrate that this approach achieves an accuracy of 98.21% in identifying four types of rice diseases, providing a reliable method for rice disease recognition research.

Eff-PCNet: An Efficient Pure CNN Network for Medical Image Classification

With the development of deep learning, convolutional neural networks (CNNs) and Transformer-based methods have become key techniques for medical image classification tasks. However, many current neural network models have problems such as high complexity, a large number of parameters, and large model sizes; such models obtain higher classification accuracy at the expense of lightweight networks. Moreover, such larger-scale models pose a great challenge for practical clinical applications. Meanwhile, Transformer and multi-layer perceptron (MLP) methods have some shortcomings in terms of local modeling capability and high model complexity, and need to be used on larger datasets to show good performance. This makes it difficult to utilize these networks in clinical medicine. Based on this, we propose a lightweight and efficient pure CNN network for medical image classification (Eff-PCNet). On the one hand, we propose a multi-branch multi-scale CNN (M2C) module, which divides the feature map into four parallel branches along the channel dimensions by a certain scale factor and carries out a deep convolution operation using different scale convolution kernels, and this multi-branch multi-scale operation effectively replaces the large kernel convolution. This multi-branch multi-scale operation effectively replaces the large kernel convolution. It reduces the computational cost of the module while fusing the feature information between different channels and thus obtains richer feature information. Finally, the four feature maps are then spliced along the channel dimensions to fuse the multi-scale and multi-dimensional feature information. On the other hand, we introduce the structural reparameterization technique and propose the structural reparameterized CNN (Rep-C) module. Specifically, it utilizes multiple linear operators to generate different feature maps during the training process and fuses all the participants into one through parameter fusion to achieve fast inference while providing a more effective solution for feature reuse. A number of experimental results show that our Eff-PCNet performs better than current methods based on CNN, Transformer, and MLP in the classification of three publicly available medical image datasets. Among them, we achieve 87.4% Acc on the HAM10000 dataset, 91.06% Acc on the SkinCancer dataset, and 97.03% Acc on the Chest-Xray dataset. Meanwhile, our approach achieves a better trade-off between the number of parameters; computation; and other performance metrics as well.

Oil well production prediction based on CNN-LSTM model with self-attention mechanism

To overcome the shortcomings in current study of oil well production prediction, we propose a combined model (CNN-LSTM-SA) with the convolutional neural network (CNN), the long short-term memory (LSTM) neural network and the self-attention mechanism (SA). The CNN-LSTM-SA model consists of five parts: input layer, CNN module, LSTM layer, self-attention layer and output layer. In this model, CNN is used to extract the spatiotemporal features of the input data, LSTM is used to extract the correlation information, and SA is used to capture the internal correlation. Compared with the traditional machine learning methods, such as linear regression (LR), support vector machine (SVM), random forest (RF), XGBoost and back propagation (BP) neural network; and deep learning methods, such as LSTM, LSTM-SA and CNN-LSTM, the CNN-LSTM-SA model can extract the spatial-temporal features that are hidden in oil well production data more comprehensively. It is enable to mine the internal correlation in oil well production data more precisely, thereby improving the accuracy of oil well production prediction. More specifically, among the existing methods, the CNN-LSTM-SA model achieves the best performance in terms of adaptation to the basic trend of oil well production and the prediction of specific values of oil well production.

Anomaly Detection Algorithm for Heterogeneous Wireless Networks Based on Cascaded Convolutional Neural Networks

As the popularity of wireless networks deepens, the diversity of device types and hardware environments makes network data take on heterogeneous forms while the threat of malicious attacks from outside can prevent ordinary methods from mining information from abnormal data. In view of this, the research will be devoted to the feature processing of the anomalous data itself, and the convolutional operation of the anomalous information by the convolutional neural network (CNN). This is to extract the internal information. In the first step of the cascaded CNNs, the dimensions of the anomaly data will be processed, the anomaly data will be sorted under the concept of relevance grouping, and then the sorted results will be added to the convolution and pooling. The performance test uses three datasets with different feature capacities as the attack sources, and the results show a 13.22% improvement in information mining performance compared to the standard CNN. The extended CNN step will perform feature identification for homologous or similar network threats, with feature expansion within the convolutional layer first, and then pooling to reduce the computational cost. The test results show that when the maximum value domain of linear expansion is 2, the model has the best feature recognition performance, fluctuating around 85%; The model comparison test results show that the accuracy of the extended CNN is higher than that of the standard CNN, and the model stability is better than that of the back propagation (BP) neural network. This indicates that the cascaded CNN dual module can mine for the data itself, thus ignoring the risk unknowns, and this connected CNN has some practical significance. The proposed cascaded CNN module applies advanced neural network technology to identify internal and external risk data. The research content has important reference value for the security management of IoT systems.

Development of a Human-Robot Interaction System for Industrial Applications

The use of robots is increasing day by day. In this study, it was aimed to develop manufacturing-assistant robot software for small production plants involving non-mass production. The main purpose of this study is to eliminate the difficulty of recruiting an expert robot operator thanks to the developed software and to facilitate the use of robots for non-experts. The developed software consists of three parts: the convolutional neural network (CNN), process selection-trajectory generation, and trajectory regulation modules. Before the operations in these modules are executed, operators record the desired process and the trajectory of the process in the video by hand gestures and index finger. Then recorded video is separated into images. The separated images are classified by the CNN module and the positions of landmarks (joint and fingernail of index finger) were calculated by the same module and using the images. Eight different pre-trained CNN structures were tested in the CNN module, and the best result Xception structure (test loss = 0.0051) was used. The desired process was determined and the trajectory of the process was created with the CNN output data. The connection of the generated trajectory with the object was detected by the trajectory regulation module, and unnecessary trajectory parts were cleaned. Regulated trajectory and desired tasks such as welding or sealing were simulated via an industrial robot in a simulation environment. As a result, an industrial robot could be programmed by non-expert operators for companies whose production line is not standard by using the developed software.

Shallow-Guided Transformer for Semantic Segmentation of Hyperspectral Remote Sensing Imagery

Convolutional neural networks (CNNs) have achieved great progress in the classification of surface objects with hyperspectral data, but due to the limitations of convolutional operations, CNNs cannot effectively interact with contextual information. Transformer succeeds in solving this problem, and thus has been widely used to classify hyperspectral surface objects in recent years. However, the huge computational load of Transformer poses a challenge in hyperspectral semantic segmentation tasks. In addition, the use of single Transformer discards the local correlation, making it ineffective for remote sensing tasks with small datasets. Therefore, we propose a new Transformer layered architecture that combines Transformer with CNN, adopts a feature dimensionality reduction module and a Transformer-style CNN module to extract shallow features and construct texture constraints, and employs the original Transformer Encoder to extract deep features. Furthermore, we also designed a simple Decoder to process shallow spatial detail information and deep semantic features separately. Experimental results based on three publicly available hyperspectral datasets show that our proposed method has significant advantages compared with other traditional CNN, Transformer-type models.

Intelligent Reception of Frequency Hopping Signals Based on CVDP

The frequency hopping communication systems have been widely used in anti-jamming communication due to their anti-interception and anti-interference capabilities. With the increasingly complex electromagnetic environment, the transmitter of the frequency hopping communication system adopts more flexible frequency hopping patterns to cope with interference. Therefore, to address the problem of traditional frequency hopping receiver systems being unable to receive properly due to intelligent frequency hopping sequence decisions made at the transmitter, we design a CVDP (CNN_VIT_Dual Multi-head Probsparse Self-attention) network based on a convolutional neural network (CNN) module and Transformer Encoder module to realize the intelligent reception of frequency hopping signals. The designed CNN module extracts the global features and spatial information of the time-frequency spectrograms and serves as the input of the VIT (Vision Transformer) network, which not only solves the problem that the VIT network is difficult to converge on small data sets but also improves the stability and robustness of the VIT network. In the Transformer Encoder module of the VIT network, a dual multi-head self-attention mechanism is used to extract critical time-frequency features and a Probsparse self-attention idea is introduced to reduce the computational complexity. Simulation results on hopping frequency estimation and received BER(Bit Error Rate) of CVDP networks in different interference and channel environments show that when a hopping sequence is unknown, CVDP-based intelligent hopping receiver systems can achieve nearly the same receiver performance as conventional reception systems in which a hopping sequence is known.

An Intelligent Detection Approach for Smoking Behavior

Smoking in public places not only causes potential harm to the health of oneself and others, but also causes hidden dangers such as fires. Therefore, for health and safety considerations, a detection model is designed based on deep learning for places where smoking is prohibited, such as airports, gas stations, and chemical warehouses, that can quickly detect and warn smoking behavior. In the model, a convolutional neural network is used to process the input frames of the video stream which are captured by the camera. After image feature extraction, feature fusion, target classification and target positioning, the position of the cigarette butt is located, and smoking behavior is determined. Common target detection algorithms are not ideal for small target objects, and the detection speed needs to be improved. A series of designed convolutional neural network modules not only reduce the amount of model calculations, speed up the deduction, and meet real-time requirements, but also improve the detection accuracy of small target objects (cigarette butts).

Electromagnetic Imaging for Buried Conductors Using Deep Convolutional Neural Networks

In the past, many conventional algorithms, such as self-adaptive dynamic differential evolution and asynchronous particle swarm optimization, were used to reconstruct buried objects in the frequency domain; these were unfortunately time-consuming during the iterative, repeated computing process of the scattered field. Consequently, we propose an innovative deep convolutional neural network approach to solve the electromagnetic inverse scattering problem for buried conductors in this paper. Different shapes of conductors are buried in one half-space and the electromagnetic wave from the other half-space is incident. The shape of the conductor can be reconstructed promptly by inputting the received scattered fields measured from the upper half-space into the deep convolutional neural network module, which avoids the computational complexity of Green’s function for training. Numerical results show that the root mean square error for differently shaped—circular, elliptical, arrow, peanut, four-petal, and three-petal—reconstructed images are, respectively, 2.95%, 3.11%, 17.81%, 15.10%, 14.14%, and 15.24%. Briefly speaking, not only can circular and elliptical buried conductors be reconstructed; some irregular shapes can be reconstructed well. On the contrary, the reconstruction result by U-Net for buried objects is worse since it is not able to obtain a good preliminary image by processing only the upper scattered field—that is, rather than the full space. In other words, our proposed deep convolutional neural network can efficiently solve the electromagnetic inverse scattering problem of buried conductors and provide a novel method for the microwave imaging of the buried conductors. This is the first successful attempt at using deep convolutional neural networks for buried conductors in the frequency domain, which may be useful for practical applications in various fields such as the medical, military, or industrial fields, including magnetic resonance imaging, mine detection and clearance, non-destructive testing, gas or wire pipeline detection, etc.

The Human Continuity Activity Semisupervised Recognizing Model for Multiview IoT Network

With advances in spatial-temporal internet of things (IoT) technologies, human activity recognition (HAR) has played a major role in human safety and medical health. Recently, most works focus on activity deep feature extraction, offering promising alternatives to manually engineered features. However, how to extract the effective and distinguishable continuity activity features and meanwhile reduce the heavy dependence on labels still remains the key challenge for human activity recognition. This paper proposes the human continuity activity semi-supervised recognizing method in multi-view IoT network scenarios. Our innovation combines supervised activity feature extraction with unsupervised encoder-decoder modules, which can capture continuity activity features from sensor data streams. To be more specific, our work applies convolutional neural network (CNN) to capture the local dependence of sensor data and designs an encoder-decoder architecture to extract temporal features in an unsupervised manner. Then we fuse these two features to recognize activities and train them with manual labels, thereby refining the temporal feature extraction and training CNN module. Experiments on five public datasets demonstrate the exceptional performance of our proposed method, which can achieve a higher recognition accuracy on almost all the datasets and is more robust and adaptive among different datasets.