Multi-scale Convolutional Network Research Articles

Research on Bearing Remaining Useful Life Anti-noise Prediction Based on Fusion of Color-Grayscale Time-Frequency Features

Abstract In contemporary industrial processes, vibration signals collected from bearings often contain significant noise, challenging the efficacy of conventional predictive models in extracting critical degradation features and accurately predicting the remaining useful life (RUL) of bearings. Addressing these challenges, this paper introduces a novel method for predicting bearing RUL under noisy conditions, leveraging a dual-branch multi-scale convolutional attention network (DMCSA) integrated with a dense residual feature fusion network (DRF). Initially, the method applies continuous wavelet trans-form (CWT) to vibration signals to extract color time-frequency image data, followed by grayscale processing to construct a comprehensive color-grayscale time-frequency image dataset, thereby augmenting the model's input features. Enhanced channel and spatial attention mechanisms, combined with multi-scale convolutions, facilitate supe-rior feature extraction and selection. The model's resilience to noise is fortified by in-corporating noise into the training dataset. Subsequently, selected color-gray time-frequency features undergo fusion and relearning through the DRF framework at the model's backend. The crayfish optimization algorithm (COA) is deployed for the astute determination of the model's critical hyperparameters. The proposed DMCSA-DRF model is then applied to predict the health indicator (MSCA-DRF-HI) of the test dataset, culminating in the accurate prediction of the bearings' RUL. Validation experiments demonstrate that our method surpasses comparative models in prediction accuracy un-der diverse noise interferences, signifying a substantial advancement in predictive performance.

Coal slurry foam image enhancement based on multiscale convolutional network

Coal slurry foam image enhancement based on multiscale convolutional network

Cross-Attention Enhanced Pyramid Multi-Scale Networks for Sensor-Based Human Activity Recognition.

Human Activity Recognition (HAR) has recently attracted widespread attention, with the effective application of this technology helping people in areas such as healthcare, smart homes, and gait analysis. Deep learning methods have shown remarkable performance in HAR. A pivotal challenge is the trade-off between recognition accuracy and computational efficiency, especially in resource-constrained mobile devices. This challenge necessitates the development of models that enhance feature representation capabilities without imposing additional computational burdens. Addressing this, we introduce a novel HAR model leveraging deep learning, ingeniously designed to navigate the accuracy-efficiency trade-off. The model comprises two innovative modules: 1) Pyramid Multi-scale Convolutional Network (PMCN), which is designed with a symmetric structure and is capable of obtaining a rich receptive field at a finer level through its multiscale representation capability; 2) Cross-Attention Mechanism, which establishes interrelationships among sensor dimensions, temporal dimensions, and channel dimensions, and effectively enhances useful information while suppressing irrelevant data. The proposed model is rigorously evaluated across four diverse datasets: UCI, WISDM, PAMAP2, and OPPORTUNITY. Additional ablation and comparative studies are conducted to comprehensively assess the performance of the model. Experimental results demonstrate that the proposed model achieves superior activity recognition accuracy while maintaining low computational overhead.

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 .

EEG emotion recognition based on the TimesNet fusion model

In recent years, emotion recognition based on electroencephalogram (EEG) has become an important research field. This paper proposes an innovative multi-scale emotion recognition method (MS-ERM), which is based on a deep learning model. First, we divide the EEG signal into time windows of 0.5 s in different frequency bands to extract the differential entropy feature and embed the feature into the brain electrode map to express spatial information. Then, the features of each segment are used as input to the new deep learning model (MS-TimesNet). The model combines multi-scale convolution and TimesNet network to effectively extract dynamic time features, cross-channel spatial features, and complex time features in 2D space. Through extensive tests on the DEAP dataset, we prove that this method is superior to existing methods in terms of sentiment classification performance. In the arousal and valence classification, the average classification accuracy of subject-dependent tests reached 91.31% and 90.45%, respectively, while in subject-independent tests, the average classification accuracy was 86.66% and 85.40%, respectively. Code is available at this repository: https://github.com/hyao0827/MS-ERM.

HEVC Video Steganalysis Based on PU Maps and Multi-Scale Convolutional Residual Network

HEVC Video Steganalysis Based on PU Maps and Multi-Scale Convolutional Residual Network

Multi-scale Convolutional Feature Fusion Network Based on Attention Mechanism for IoT Traffic Classification

AbstractThe Internet of Things (IoT) has been extensively utilized in domains such as smart homes, healthcare, and other industries. With the exponential growth of Internet of Things (IoT) devices, they have become prime targets for malicious cyber-attacks. Effective classification of IoT traffic is, therefore, imperative to enable robust intrusion detection systems. However, IoT traffic data contain intricate spatial relationships and topological information, which traditional methods for traffic identification lack the capability to fully extract features and capture crucial characteristics. We propose a multi-scale convolutional feature fusion network augmented with a Convolutional Block Attention Module (MCF-CBAM) for accurate IoT traffic classification. The network incorporates three critical innovations: (1) Parallel convolution extracts multi-scale spatial features from traffic data. The 1 × 1 convolution operation reduces the amount of parameters and calculations of the network, thereby improving work efficiency. (2) The attention module suppresses less informative features while highlighting the most discriminative ones, enabling focused learning on decisive features. (3) Cross-scale connections with channel jumps reuse features from prior layers to enhance generalization. We evaluate the method extensively on three widely adopted public datasets. Quantitative results demonstrate MCF-CBAM establishes new state-of-the-art performance benchmarks for IoT traffic classification, surpassing existing methods by a significant margin. Qualitative visualizations of the learned attention weights provide intuitive insights into how the network automatically discovers the most decisive spatial features for identification. With its strong empirical performance and interpretable attention mechanisms, this work presents a promising deep learning solution to augment real-world IoT intrusion detection systems against growing cybersecurity threats.

EEG-FRM: a neural network based familiar and unfamiliar face EEG recognition method.

Recognizing familiar faces holds great value in various fields such as medicine, criminal investigation, and lie detection. In this paper, we designed a Complex Trial Protocol-based familiar and unfamiliar face recognition experiment that using self-face information, and collected EEG data from 147 subjects. A novel neural network-based method, the EEG-based Face Recognition Model (EEG-FRM), is proposed in this paper for cross-subject familiar/unfamiliar face recognition, which combines a multi-scale convolutional classification network with the maximum probability mechanism to realize individual face recognition. The multi-scale convolutional neural network extracts temporal information and spatial features from the EEG data, the attention module and supervised contrastive learning module are employed to promote the classification performance. Experimental results on the dataset reveal that familiar face stimuli could evoke significant P300 responses, mainly concentrated in the parietal lobe and nearby regions. Our proposed model achieved impressive results, with a balanced accuracy of 85.64%, a true positive rate of 73.23%, and a false positive rate of 1.96% on the collected dataset, outperforming other compared methods. The experimental results demonstrate the effectiveness and superiority of our proposed model.

ECT: Fine-grained edge detection with learned cause tokens

In this study, we tackle the challenging fine-grained edge detection task, which refers to predicting specific edges caused by reflectance, illumination, normal, and depth changes, respectively. Prior methods exploit multi-scale convolutional networks, which are limited in three aspects: (1) Convolutions are local operators while identifying the cause of edge formation requires looking at far away pixels. (2) Priors specific to edge cause are fixed in prediction heads. (3) Using separate networks for generic and fine-grained edge detection, and the constraint between them may be violated. To address these three issues, we propose a two-stage transformer-based network sequentially predicting generic edges and fine-grained edges, which has a global receptive field thanks to the attention mechanism. The prior knowledge of edge causes is formulated as four learnable cause tokens in a cause-aware decoder design. Furthermore, to encourage the consistency between generic edges and fine-grained edges, an edge aggregation and alignment loss is exploited. We evaluate our method on the public benchmark BSDS-RIND and several newly derived benchmarks, and achieve new state-of-the-art results. Our code, data, and models are publicly available at https://github.com/Daniellli/ECT.git.

12-Lead ECG signal classification for detecting ECG arrhythmia via an information bottleneck-based multi-scale network

The 12-lead electrocardiogram (ECG) is a reliable diagnostic tool for detecting and treating severe cardiovascular conditions like arrhythmia and heart attack. Deep neural networks (DNNs) have achieved higher accuracy in recent years than traditional ECG signal classification task methods. Convolutional neural network (CNN) and Transformer are the two mainstream architectures of DNN, respectively good at extracting local and global features from input data. This paper proposes the multi-scale convolutional Transformer network (MCTnet), an efficient combination of Transformer encoder and CNN for ECG signal classification. MCTnet utilizes the advantages of CNN and self-attention mechanisms to capture potential features in ECG signal accurately. The dual-branch Transformer encoder extracts different-scale feature representations, enabling the capture of both local and global information. Additionally, an information bottleneck method eliminates redundant information and enhances task-relevant information in the learned representations. To evaluate the performance of MCTnet, comprehensive experiments are conducted on three commonly used ECG datasets. The results demonstrate that MCTnet outperforms current deep learning-based models, highlighting its effectiveness in ECG signal classification. It also shows that the performance of the model can be effectively improved by utilizing multi-scale representation learning and information bottleneck.

Multiscale convolutional conditional domain adversarial network with channel attention for unsupervised bearing fault diagnosis

Unsupervised cross-domain fault diagnosis is an effective technical way to realize the engineering application of bearing fault diagnosis methods. However, there are still two problems that need to be resolved. First, the importance of fault features at different scales is generally not consistent. There is redundant information in the fault features. Second, most methods mainly study how to lessen the marginal distribution difference in source and target domains while ignoring their class information. When the data distribution contains complex multimodal structure, this may lead to failure to capture the multimodal structure. To address the above problems, a multiscale channel attention conditional domain adversarial network is proposed. First, a new channel attention module is designed to assign different weights to different channels, which can highlight valuable features and stamp out superfluous features. Then, conditional domain adversarial is used to fully capture the multimodal structure through cross-covariance dependencies between features and classes. Our method’s capability is validated by diagnose results on public data sets and self-built data sets.

Multiscale Convolutional Mask Network for Hyperspectral Unmixing

Multiscale Convolutional Mask Network for Hyperspectral Unmixing

Multi-Scale Convolutional Echo State Network With an Effective Pre-Training Strategy for Solar Irradiance Forecasting

Multi-Scale Convolutional Echo State Network With an Effective Pre-Training Strategy for Solar Irradiance Forecasting

A Fast Evolutionary Knowledge Transfer Search for Multiscale Deep Neural Architecture.

The emergence of neural architecture search (NAS) algorithms has removed the constraints on manually designed neural network architectures, so that neural network development no longer requires extensive professional knowledge, trial and error. However, the extremely high computational cost limits the development of NAS algorithms. In this article, in order to reduce computational costs and to improve the efficiency and effectiveness of evolutionary NAS (ENAS) is investigated. In this article, we present a fast ENAS framework for multiscale convolutional networks based on evolutionary knowledge transfer search (EKTS). This framework is novel, in that it combines global optimization methods with local optimization methods for search, and searches a multiscale network architecture. In this article, evolutionary computation is used as a global optimization algorithm with high robustness and wide applicability for searching neural architectures. At the same time, for fast search, we combine knowledge transfer and local fast learning to improve the search speed. In addition, we explore a multiscale gray-box structure. This gray box structure combines the Bandelet transform with convolution to improve network approximation, learning, and generalization. Finally, we compare the architectures with more than 40 different neural architectures, and the results confirmed its effectiveness.

Transformer encoder with multiscale deep learning for pain classification using physiological signals.

Pain, a pervasive global health concern, affects a large segment of population worldwide. Accurate pain assessment remains a challenge due to the limitations of conventional self-report scales, which often yield inconsistent results and are susceptible to bias. Recognizing this gap, our study introduces PainAttnNet, a novel deep-learning model designed for precise pain intensity classification using physiological signals. We investigate whether PainAttnNet would outperform existing models in capturing temporal dependencies. The model integrates multiscale convolutional networks, squeeze-and-excitation residual networks, and a transformer encoder block. This integration is pivotal for extracting robust features across multiple time windows, emphasizing feature interdependencies, and enhancing temporal dependency analysis. Evaluation of PainAttnNet on the BioVid heat pain dataset confirm the model's superior performance over the existing models. The results establish PainAttnNet as a promising tool for automating and refining pain assessments. Our research not only introduces a novel computational approach but also sets the stage for more individualized and accurate pain assessment and management in the future.

A Multiscale Deep Encoder–Decoder with Phase Congruency Algorithm Based on Deep Learning for Improving Diagnostic Ultrasound Image Quality

Ultrasound imaging is widely used as a noninvasive lesion detection method in diagnostic medicine. Improving the quality of these ultrasound images is very important for accurate diagnosis, and deep learning-based algorithms have gained significant attention. This study proposes a multiscale deep encoder–decoder with phase congruency (MSDEPC) algorithm based on deep learning to improve the quality of diagnostic ultrasound images. The MSDEPC algorithm included low-resolution (LR) images and edges as inputs and constructed a multiscale convolution and deconvolution network. Simulations were conducted using the Field 2 program, and data from real experimental research were obtained using five clinical datasets containing images of the carotid artery, liver hemangiomas, breast malignancy, thyroid carcinomas, and obstetric nuchal translucency. LR images, bicubic interpolation, and super-resolution convolutional neural networks (SRCNNs) were modeled as comparison groups. Through visual assessment, the image processed using the MSDEPC was the clearest, and the lesions were clearly distinguished. The structural similarity index metric (SSIM) value of the simulated ultrasound image using the MSDEPC algorithm improved by approximately 38.84% compared to LR. In addition, the peak signal-to-noise ratio (PSNR) and SSIM values of clinical ultrasound images using the MSDEPC algorithm improved by approximately 2.33 times and 88.58%, respectively, compared to LR. In conclusion, the MSDEPC algorithm is expected to significantly improve the spatial resolution of ultrasound images.

Predictive Recognition of DNA-binding Proteins Based on Pre-trained Language Model BERT.

Identifying proteins is crucial for disease diagnosis and treatment. With the increase of known proteins, large-scale batch predictions are essential. However, traditional biological experiments being time-consuming and expensive are difficult to accomplish this task efficiently. Nevertheless, deep learning algorithms based on big data analysis have manifested potential in this aspect. In recent years, language representation models, especially BERT, have made significant advancements in natural language processing. In this paper, using three protein segmentation methods and three encoder numbers, nine BERT models with different sizes are constructed to predict whether known proteins are DNA-binding proteins or not. Furthermore, based on the concept of protein motifs, multi-scale convolutional networks are fused into the models to extract the local features of DNA-binding proteins. Finally, we find that the larger the number of encoders, the better the model predictions under the condition of considering each amino acid in the protein as a word. Our proposed algorithm achieves 81.88% sensitivity and 0.39 MCC value on the test set. Furthermore, it achieves 62.41% accuracy on the independent test set PDB2272. It is evident that our proposed method can be a tool to assist in the identification of DNA-binding proteins.

RUL prediction of rolling bearings across working conditions based on multi-scale convolutional parallel memory domain adaptation network

Rolling bearings are widely used in mechanical equipment, effectively determining the failure time of rolling bearings is particularly significant to ensure the safe performance of mechanical equipment. However, in industrial scenarios, the machine mainly works in the normal state for a long time, it is difficult to accumulate the same distribution of the whole life data, but the use of different distribution of data for forecasting will reduce the performance of deep learning-based prediction methods. Therefore, in order to tackle this problem, a multi-scale convolutional parallel memory domain adaptation network is investigated to forecast the residual useful life (RUL) of bearings across working conditions. Firstly, a new characteristic extractor—multi-scale convolutional parallel memory network is designed to extract spatial and temporal characteristics of bearing degradation data. At the same time, in order to minimize the distribution difference between source domain and target domain, a temporal-spatial feature alignment strategy is proposed to obtain domain invariable characteristics by combining maximum mean difference and domain adversarial learning. Finally, the availability of the proposed approach is verified using two rolling bearing data sets. The results reveal that it can efficiently forecast the RUL of rolling bearings across working conditions.

Multiscale convolutional neural-based transformer network for time series prediction

Multiscale convolutional neural-based transformer network for time series prediction

Novel imbalanced subdomain adaption multiscale convolutional network for cross-domain unsupervised fault diagnosis of rolling bearings

Data on the vibration signals collected from rolling bearings mostly belongs to health conditions, leading to an imbalanced data distribution. In addition, frequent switching of operating conditions results in unlabeled data collected under a specific working condition. This paper proposes a novel network for cross-domain unsupervised fault diagnosis of rolling bearings considering the imbalanced data to address these challenges. First, a multiscale parallel features extraction is developed, which can fully mine the rich high-level feature representation of various fault types from the original data and has a high value for fault identification. Second, a squeeze-and-excitation attention mechanism is constructed to enhance features conducive to model classification and suppress redundant features. Finally, a new loss function is proposed to optimize the model, which can accurately classify imbalanced source domain and easily align related subdomains of two domains. The proposed method was validated on multiple unsupervised cross-domain diagnostic tasks on two bearing datasets. Experimental results manifest that the proposed method has stable generalization performance and excellent robustness.