Dilated Convolutional Neural Network Research Articles

Multiscale dilated convolutional neural network for Atrial Fibrillation detection.

Atrial Fibrillation (AF), a type of heart arrhythmia, becomes more common with aging and is associated with an increased risk of stroke and mortality. In light of the urgent need for effective automated AF monitoring, existing methods often fall short in balancing accuracy and computational efficiency. To address this issue, we introduce a framework based on Multi-Scale Dilated Convolution (AF-MSDC), aimed at achieving precise predictions with low cost and high efficiency. By integrating Multi-Scale Dilated Convolution (MSDC) modules, our model is capable of extracting features from electrocardiogram (ECG) datasets across various scales, thus achieving an optimal balance between precision and computational savings. We have developed three MSDC modules to construct the AF-MSDC framework and assessed its performance on renowned datasets, including the MIT-BIH Atrial Fibrillation Database and Physionet Challenge 2017. Empirical results unequivocally demonstrate that our technique surpasses existing state-of-the-art (SOTA) methods in the AF detection domain. Specifically, our model, with only a quarter of the parameters of a Residual Network (ResNet), achieved an impressive sensitivity of 99.45%, specificity of 99.64% (on the MIT-BIH AFDB dataset), and an [Formula: see text] score of 85.63% (on the Physionet Challenge 2017 AFDB dataset). This high efficiency makes our model particularly suitable for integration into wearable ECG devices powered by edge computing frameworks. Moreover, this innovative approach offers new possibilities for the early diagnosis of AF in clinical applications, potentially improving patient quality of life and reducing healthcare costs.

LSTM-QDCNN: long short-term memory and quantum dilated convolutional neural network enabled occlusion percentage prediction

ABSTRACT Occlusion detection is vital in longer video sequences of real-time scenarios, as deformable objects generally occlude themselves or be occluded by exterior objects eventually. Occlusion can affect accuracy of the entire model significantly, because optical flow may reveal random attributes in occluded as well as neighbourhood areas. Occlusion is a challenging issue in real-world visual object tracking. Several techniques learn imprecise appearance of target while it becomes occluded by another object in the scene. Occlusion is a complicated issue that causes target loss in tracking procedure. Here, long short-term memory fused quantum dilated convolutional neural network (LSTM-QDCNN) is presented for occlusion percentage prediction. Initially, video frames are extracted from considered input video. The extracted frame is pre-processed by adaptive fuzzy filtering, and object localisation is accomplished by employing sparse fuzzy c-means with local optimal oriented pattern feature. Thereafter, occlusion category detection is conducted by deep Q-network. Lastly, occlusion percentage prediction is done by utilising LSTM-QDCNN. Additionally, LSTM-QDCNN is designed by merging deep LSTM (DLSTM) with quantum dilated convolutional neural network (QDCNN). Furthermore LSTM-QDCNN achieved maximum multiple-object tracking precision of 0.892 and minimum tracking distance of 3.714. In addition, LSTM-QDCNN achieved tracking number about 18.

Predicting Tool Wear with ParaCRN-AMResNet: A Hybrid Deep Learning Approach

In the manufacturing sector, tool wear substantially affects product quality and production efficiency. While traditional sequential deep learning models can handle time-series tasks, their neglect of complex temporal relationships in time-series data often leads to errors accumulating in continuous predictions, which reduces their forecasting accuracy for tool wear. For addressing these limitations, the parallel convolutional and recurrent neural networks with attention-modulated residual learning (ParaCRN-AMResNet) model is introduced. Compared with conventional deep learning models, ParaCRN-AMResNet markedly enhances the efficiency and precision of feature extraction from time-series data through its innovative parallel architecture. The model adeptly combines dilated convolution neural network and bidirectional gated recurrent units, effectively addressing distance dependencies and enriching the quantity and dimensions of extracted features. The strength of ParaCRN-AMResNet lies in its refined ability to capture the complex dynamics of time-series data, significantly boosting the model’s accuracy and generalization capability. The model’s efficacy was validated through comprehensive milling experiments and vibration signal analyses, showcasing ParaCRN-AMResNet’s superior performance. In evaluation metrics, the model achieved a MAE of 2.6015, MSE of 15.1921, R2 of 0.9897, and MAPE of 2.7997%, conclusively proving its efficiency and accuracy in the precise prediction of tool wear.

TriConvUNeXt: A Pure CNN-Based Lightweight Symmetrical Network for Biomedical Image Segmentation.

Biomedical image segmentation is essential in clinical practices, offering critical insights for accurate diagnosis and strategic treatment approaches. Nowadays, self-attention-based networks have achieved competitive performance in both natural language processing and computer vision, but the computational cost has reduced their popularity in practical applications. The recent study of Convolutional Neural Network (CNN) explores linear functions within modified CNN layer demonstrating pure CNN-based networks can still achieve competitive results against Vision Transformer (ViT) in biomedical image segmentation, with fewer parameters. The modified CNN, i.e., Depthwise CNN, however, leaves the features cleaved off in the channel dimension and prevents the extraction of features in the perspective of channel interaction. To effectively further explore the feature learning power of modified CNN with biomedical image segmentation, we design a lightweight multi-convolutional multi-scale convolutional network block (MSConvNeXt) for U-shape symmetrical network. Specifically, a network block consisting of a depthwise CNN, a deformable CNN, and a dilated CNN is proposed to capture semantic feature information effectively while with low computational cost. Furthermore, channel shuffling operation is proposed to dynamically promote an efficient feature fusion among the feature maps. The network block hereby is properly deployed within U-shape symmetrical encoder-decoder style network, named TriConvUNeXt. The proposed network is validated on a public benchmark dataset with a comprehensive evaluation in both computational cost and segmentation performance against 13 baseline methods. Specifically, TriConvUNeXt achieves 1% higher than UNet and TransUNet in Dice-Coefficient while 81% and 97% lower in computational cost. The implementation of TriConvUNeXt is made publicly accessible via https://github.com/ziyangwang007/TriConvUNeXt .

Sentiment analysis in social internet of things using contextual representations and dilated convolution neural network

Sentiment analysis in social internet of things using contextual representations and dilated convolution neural network

Blind recognition of channel codes based on a multiscale dilated convolution neural network

Channel coding is an essential part of communication. In a non-cooperative communication system, the receiver has to recover the signal without any priori knowledge. Therefore, it is important to recognize channel code types and parameters. At present, only a few studies focus on channel code type recognition. In this paper, a multiscale dilated convolution neural network (MSDCNN) is proposed to achieve fine-grained recognition of seven commonly used channel code types. MSDCNN has three multiscale modules. Based on the characteristics of each channel code type, the first two identical modules are designed to extract the overall features of codewords with unit convolution and three parallel regular convolution layers. To lower complexity, the third module uses dilated convolution instead of regular convolution to extract features between codewords. Simulation results show that MSDCNN has higher average recognition accuracy than other neural networks. More importantly, MSDCNN lowers complexity, making it more suitable for deployment so that it can be applied in practical non-cooperative receivers.

SADCNN-ORBM: a hybrid deep learning model based citrus disease detection and classification

Citrus disease has a significant influence on agricultural productivity these days, so technology based on artificial intelligence has been developed for creating computer vision (CV) models. By spotting disease in its early stages and enabling necessary productivity actions, CV in agriculture improves the production of agricultural goods. In this paper, we developed a CV-based citrus disease detection model called the self-attention dilated convolutional neural network optimized restricted Boltzmann machine (SADCNN-ORBM) model, which consists of two crucial parts: a SADCNN for disease segmentation and an ORBM optimized by the self-adaptive coati optimization (SACO) algorithm to improve the classification performance of diseases, which successfully divides the disease type into three groups: anthracnose, melanose, and brown spot. Numerous feature sets, such as texture features, three-channel red, green, blue (RGB) features, local binary pattern (LBP) features, and speeded-up robust features (SURF) features, are combined and given as input into the classification layer in the proposed model. We compare our proposed model's performance with existing methods by using several evaluation metrics. The findings demonstrate the SADCNN-ORBM model's superiority in precisely recognizing and classifying citrus illnesses, outperforming all available techniques.

A Data-Driven Autonomous Assessment Framework for Education Quality Based on Multiscale Deep Learning

Nowadays, computational intelligence-assisted autonomous assessment of education quality has become a more and more general concern in the area of smart education management. As education quality assessment is a complicated process with multiple heterogeneous factors, it remains challenging to make effective assessment using simple information modality and criteria. To deal with this issue, this paper introduces multiscale deep learning, and proposes a novel data-driven autonomous assessment framework for education quality. In particular, four aspects of heterogeneous indicators are selected as the basic indexes and a dilated convolutional neural network structure is formulated to perform multiscale feature extraction. Then, the structural equation is adopted to make multivariate characterization and output the final assessment results. At last, some simulations are carried out on realistic education operation data to evaluate the efficiency of the proposed autonomous assessment framework. Two aspects of findings can be deduced from the results. For one thing, multisource fusion of multiple indicators well makes sense to autonomous education assessment. For another, multiscale deep learning can provide some beneficial promotion to assessment efficiency.

The graft‐versus‐host problem for data‐driven gravity‐wave parameterizations in a one‐dimensional quasibiennial oscillation model

AbstractTwo key challenges in the development of data‐driven gravity‐wave parameterizations are generalization, how to ensure that a data‐driven scheme trained on the present‐day climate will continue to work in a new climate regime, and calibration, how to account for biases in the “host” climate model. Both problems depend fundamentally on the response to out‐of‐sample inputs compared with the training dataset, and are often conflicting. The ability to generalize to new climate regimes often goes hand in hand with sensitivity to model biases. To probe these challenges, we employ a one‐dimensional (1D) quasibiennial oscillation (QBO) model with a stochastic source term that represents convectively generated gravity waves in the Tropics with randomly varying strengths and spectra. We employ an array of machine‐learning models consisting of a fully connected feed‐forward neural network, a dilated convolutional neural network, an encoder–decoder, a boosted forest, and a support‐vector regression model. Our results demonstrate that data‐driven schemes trained on “observations” can be critically sensitive to model biases in the wave sources. While able to emulate accurately the stochastic source term on which they were trained, all of our schemes fail to simulate fully the expected QBO period or amplitude, even with the slightest perturbation to the wave sources. The main takeaway is that some measures will always be required to ensure the proper response to climate change and to account for model biases. We examine one approach based on the ideas of optimal transport, where the wave sources in the model are first remapped to the observed one before applying the data‐driven scheme. This approach is agnostic to the data‐driven method and guarantees that the model adheres to the observational constraints, making sure the model yields the right results for the right reasons.

Improved polarization scattering imaging using local-global context polarization feature learning framework

Polarization scattering imaging (PSI) is an extremely challenging problem due to the fact that scattering media can lead to severe degradation of the object information. In this paper, we propose a novel high-performance computational method, the PSI based on a well-designed local-global context polarization feature learning (LGCPFL) framework, named PSI-LGCPFL, to efficiently retrieve the target’s information. In the LGCPFL framework, the dilated convolutional neural network and Swin Transformer are concurrently introduced to capture local polarization feature information and long-range polarization dependencies in the polarization scattering images, respectively. We in-depth investigate the relationship between the geometry of target, the scattering imaging distance, and the constituent materials and the PSI quality. The performance of our proposed PSI-LGCPFL is benchmarked and evaluated on three testing datasets established in real-life scenarios, i.e., STR24, DIS50 and MAT18, which cover untrained target’s geometry, various untrained targets lying in untrained distances between scattering medium and targets, and diverse untrained target’s materials with different background materials, respectively. The experimental results demonstrate that the proposed PSI-LGCPFL has a superior performance on retrieving the target’s information with high-generalization abilities and reasonably inferring speed, and outperforms several existing state-of-the-art methods. To our knowledge, PSI-LGCPFL is the first approach to achieve the PSI by polarization characteristics and a deep learning model with Swin Transformer. It also highlights the prospect of accurately reconstruction of remote sensing target’s information in scattering medium via using polarization information and deep learning

DFDTA-MultiAtt: Multi-attention based deep learning ensemble fusion network for drug target affinity prediction

<p>An essential step in the drug development process is the accurate detection of drug-target interactions (DTI). The importance of binding affinity values in understanding protein-ligand interactions was previously disregarded, and DTI prediction was only seen as a binary classification problem. In this regard, we introduced the DFDTA-MultiAtt model for predicting the drug target binding affinity in two stages using the structural and sequential information. The first step of the first stage involves retrieving features from sequence data using a bi-directional long short term memory (Bi-LSTM) architecture together with a multi-attention module and dilated convolutional neural network (dilated-CNN) architecture, and the second step features are learnt from structure representation once again using a dilated-CNN. To predict the binding affinity, the second stage uses an ensemble learning model. The proposed model also produces findings with a greater overall accuracy when compared to contemporary state-of-the-art methods. The model generates an enormous +0.006 concordance index (CI) score on the Davis dataset and reduces the mean square error (MSE) by 0.174 on the KIBA dataset.</p>

Advanced hybrid attention-based deep learning network with heuristic algorithm for adaptive CT and PET image fusion in lung cancer detection

Lung cancer is one of the most deadly diseases in the world. Lung cancer detection can save the patient's life. Despite being the best imaging tool in the medical sector, clinicians find it challenging to interpret and detect cancer from Computed Tomography (CT) scan data. One of the most effective ways for the diagnosis of certain malignancies like lung tumours is Positron Emission Tomography (PET) imaging. So many diagnosis models have been implemented nowadays to diagnose various diseases. Early lung cancer identification is very important for predicting the severity level of lung cancer in cancer patients. To explore the effective model, an image fusion-based detection model is proposed for lung cancer detection using an improved heuristic algorithm of the deep learning model. Firstly, the PET and CT images are gathered from the internet. Further, these two collected images are fused for further process by using the Adaptive Dilated Convolution Neural Network (AD-CNN), in which the hyperparameters are tuned by the Modified Initial Velocity-based Capuchin Search Algorithm (MIV-CapSA). Subsequently, the abnormal regions are segmented by influencing the TransUnet3+. Finally, the segmented images are fed into the Hybrid Attention-based Deep Networks (HADN) model, encompassed with Mobilenet and Shufflenet. Therefore, the effectiveness of the novel detection model is analyzed using various metrics compared with traditional approaches. At last, the outcome evinces that it aids in early basic detection to treat the patients effectively.

Meta-Transfer Metric Learning for Time Series Classification in 6G-Supported Intelligent Transportation Systems

Deep learning-based time series classification in 6G-supported Intelligent Transportation Systems (ITS) helps transport decision-making. Deep learning classifier training necessitates a large amount of labeled data for feature extraction. Labeling time series data in 6G-supported ITS is tough. Meta-learning can be used to train deep classifiers with limited data. However, in meta-learning, the tasks are frequently modeled by a low-complexity base learner. It is unable to use more complicate and powerful structures. The meta-learning-pretrained classifier can only perform new classification problems with the same number of classes. Most pre-training strategies do not prioritize enhancing the pre-training phase’s convergence rate and lowering the computational cost. Most research work aims to improve classification performance by increasing the complexity of the classification model. However, this raises computing costs. In this paper, we propose a one-dimensional Multi-Scale Dilated Convolution Neural Network time series classifier (MSDCNN). MSDCNN combines multi-scale CNN and dilated CNN. It can extract multi-scale characteristics from time series and reduce the complexity of the classifier. Furthermore, we propose a pre-training strategy, called Meta-transfer metric Learning using Scale function (MLS). MLS allows the classifier to gain experience from different tasks with various numbers of classes. Experiments show that MLS reduces pre-training computation costs during the pre-training phase. The pre-trained classifier, without using any fine tuning techniques, achieves the highest accuracy by comparing with the state-of-the-art methods. Finally, we present a case study of applying MSDCNN and MLS to detect road accidents in 6G-supported transportation systems.

Archimedes optimisation algorithm quantum dilated convolutional neural network for road extraction in remote sensing images

Roads are closely intertwined with human existence, and the process of extracting road networks has emerged as the most prominent task in remote sensing (RS). The automated road interpretation process of remote sensing images (RSI) efficiently acquires road network data at a reduced expense in comparison to the traditional visual interpretation of RSI. However the manifestation of RSI is completely distinct because of the great difference in length, width, material, and shape of road networks in dissimilar areas. Thus, the extraction of road network data in RSI is still a complex issue. In recent times, DL-based approaches have projected a famous development in image segmentation outcomes, but a lot of them still could not retain boundary data and attain high-resolution road segmentation maps while processing the RSI. Traditional convolutional neural networks (CNNs) demonstrate impressive performance in road extract tasks; however, they frequently encounter difficulties in capturing intricate details and contextual information. The study introduces a novel method, named Archimedes Optimisation Algorithm, Quantum Dilated Convolutional Neural Network for Road Extraction (AOA-QDCNNRE), to tackle the challenges encountered in remote sensing images. The AOA-QDCNNRE technique aims to generate a high-resolution road segmentation map using DL with a hyperparameter tuning process. The AOA-QDCNNRE technique primarily relies on the QDCNN model, which integrates quantum technology (QC) with dilated convolutions to augment the network's capacity to capture local as well as global contextual information. In addition, the incorporation of the dilated convolutional technique effectively enhances the receptive field without sacrificing spatial resolution, enabling the extraction of precise road features. To develop the road extraction outcomes of the QDCNN approach, the AOA-based hyperparameter tuning process can be exploited. The AOA-QDCNNRE system's simulation results can be tested on benchmark databases, and the results indicate that the AOA-QDCNNRE method surpasses recent algorithms.

FastEval Parkinsonism: an instant deep learning–assisted video-based online system for Parkinsonian motor symptom evaluation

The Motor Disorder Society’s Unified Parkinson’s Disease Rating Scale (MDS-UPDRS) is designed to assess bradykinesia, the cardinal symptoms of Parkinson’s disease (PD). However, it cannot capture the all-day variability of bradykinesia outside the clinical environment. Here, we introduce FastEval Parkinsonism (https://fastevalp.cmdm.tw/), a deep learning-driven video-based system, providing users to capture keypoints, estimate the severity, and summarize in a report. Leveraging 840 finger-tapping videos from 186 individuals (103 patients with Parkinson’s disease (PD), 24 participants with atypical parkinsonism (APD), 12 elderly with mild parkinsonism signs (MPS), and 47 healthy controls (HCs)), we employ a dilated convolution neural network with two data augmentation techniques. Our model achieves acceptable accuracies (AAC) of 88.0% and 81.5%. The frequency-intensity (FI) value of thumb-index finger distance was indicated as a pivotal hand parameter to quantify the performance. Our model also shows the usability for multi-angle videos, tested in an external database enrolling over 300 PD patients.

The Value of Large‐Scale Climatic Indices for Monthly Forecasting Severity of Widespread Flooding Using Dilated Convolutional Neural Networks

AbstractSpatially co‐occurring floods pose a threat to the resilience and recovery of the communities. Their timely forecasting plays a crucial role for increasing flood preparedness and limiting associated losses. In this study we investigated the potential of a dilated Convolutional Neural Network (dCNN) model conditioned on large‐scale climatic indices and antecedent precipitation to forecast monthly severity of widespread flooding (i.e., spatially co‐occurring floods) in Germany with 1 month lead time. The severity was estimated from 63 years of daily streamflow series as the sum of concurrent exceedances of at‐site 2‐year return periods within a given month across 172 mesoscale catchments (median area 516 km2). The model was trained individually for the whole country and three diverse hydroclimatic regions to provide insights on heterogeneity of model performance and flood drivers. Our results showed a considerable potential for forecasting widespread flood severity using dCNN especially as the length of training series increases. However, event‐based evaluation of model skill indicates large underestimation for rainfall‐generated floods during dry conditions despite overall lower severity of these events compared to the rain‐on‐snow floods. Feature attribution and wavelet coherence analyses both indicated considerable difference in the major flood drivers in three regions. While the flooding in North‐Eastern region is strongly affected by the Baltic Sea, the North‐Western region is affected more by global patterns associated with the El‐Niño activity. In the Southern region in addition to global patterns we detected the effect of the Mediterranean Sea, while antecedent precipitation plays a less important role in this region.

A Semi-Supervised Multi-Scale Arbitrary Dilated Convolution Neural Network for Pediatric Sleep Staging.

Sleep staging is essential for assessing sleep quality and diagnosing sleep disorders. However, sleep staging is a labor-intensive process, making it arduous to obtain large quantities of high-quality labeled data for automatic sleep staging. Meanwhile, most of the research on automatic sleep staging pays little attention to pediatric sleep staging. To address these challenges, we propose a semi-supervised multi-scale arbitrary dilated convolution neural network (SMADNet) for pediatric sleep staging using the scalogram with a high height-to-width ratio generated by the continuous wavelet transform (CWT) as input. To extract more extended time dimensional feature representations and adapt to scalograms with a high height-to-width ratio in SMADNet, we introduce a multi-scale arbitrary dilation convolution block (MADBlock) based on our proposed arbitrary dilated convolution (ADConv). Finally, we also utilize semi-supervised learning as the training scheme for our network in order to alleviate the reliance on labeled data. Our proposed model has achieved performance comparable to state-of-the-art supervised learning methods with 30% labels. Our model is tested on a private pediatric dataset and achieved 79% accuracy, 72% kappa, and 75% MF1. Therefore, our model demonstrates a powerful feature extraction capability and has achieved performance comparable to state-of-the-art supervised learning methods with a small number of labels.

Recursive signal denoising method for predictive maintenance of equipment by using deep learning based temporal masking

In the data driven predictive maintenance, high quality data is the premise of high accuracy diagnosis and prediction. In the industrial practice, reducing the noise is of great significance to ensure data quality. This paper proposes a recursive denoising method for manufacturing equipment signals in the data driven predictive maintenance. First, in signal decomposition method, equipment mixed signal is decomposed by temporal masking with dilated convolution neural network to generate a noise mask, which realizes signal decomposition of using recursive operation of temporal masking model. Second, in signal components recognition method, signal component features similarities are calculated, which act on the parameter regulation of signal recognition meta-learning model. The experimental results demonstrated that the proposed method effectively solves the noise reduction problem of the equipment signal. Further engineering tests of a chemical winding machine vibration signal decomposition and recognition show that the proposed method has strong adaptive performance for noise reduction.

Inter-turn short circuit and demagnetization fault diagnosis of ship PMSM based on multiscale residual dilated CNN and BiLSTM

Inter-turn short circuit (ITSC) and demagnetization of permanent magnet synchronous motors (PMSMs) can lead to serious ship accidents, timely and accurate fault diagnosis of these faults is very important. A multi-signal fusion fault diagnosis method (MD-CNN-BiLSTM) is proposed based on multi-scale residual dilated convolutional neural network (D-CNN) and bidirectional long and short-term memory (BiLSTM) for PMSM fault diagnosis. This method first takes three-phase current and vibration signals as input; uses a three-column parallel CNN structure with different scales to extract both global signal and local feature. A residual connection in the expanded CNN is then used to eliminate the problems of gradient disappearance or explosion; and finally, BiLSTM is used to further extract features and identify the fault. A 2.2 kW permanent magnet synchronous motor was used to build a fault simulation test rig. The motor stator was rewound to simulate the ITSC fault, and different sizes of permanent magnets were replaced to simulate demagnetization fault. ITSC, demagnetization and their coupled faults were simulated under 10 specific motor speeds and loads respectively. The test proved that the diagnostic accuracy of the proposed method was 4.2% higher than that of ordinary CNN and 29.06% higher than that of BiLSTM. It also had the best diagnostic effect under the noise interference of different intensities. It was verified that the proposed method has good noise interference and strong classification ability.

DCNNLFS: A Dilated Convolutional Neural Network with Late Fusion Strategy for Intelligent Classification of Gastric Histopathology Images.

Gastric cancer has a high incidence rate, significantly threatening patients' health. Gastric histopathology images can reliably diagnose related diseases. Still, the data volume of histopathology images is too large, making misdiagnosis or missed diagnosis easy. The classification model based on deep learning has made some progress on gastric histopathology images. However, traditional convolutional neural networks (CNN) generally use pooling operations, which will reduce the spatial resolution of the image, resulting in poor prediction results. The image feature in previous CNN has a poor perception of details. Therefore, we design a dilated CNN with a late fusion strategy (DCNNLFS) for gastric histopathology image classification. The DCNNLFS model utilizes dilated convolutions, enabling it to expand the receptive field. The dilated convolutions can learn the different contextual information by adjusting the dilation rate. The DCNNLFS model uses a late fusion strategy to enhance the classification ability of DCNNLFS. We run related experiments on a gastric histopathology image dataset to verify the excellence of the DCNNLFS model, where the three metrics Precision, Accuracy, and F1-Score are 0.938, 0.935, and 0.959.