Channel Dimensions Research Articles

ELA-Net: An Efficient Lightweight Attention Network for Skin Lesion Segmentation.

In clinical conditions limited by equipment, attaining lightweight skin lesion segmentation is pivotal as it facilitates the integration of the model into diverse medical devices, thereby enhancing operational efficiency. However, the lightweight design of the model may face accuracy degradation, especially when dealing with complex images such as skin lesion images with irregular regions, blurred boundaries, and oversized boundaries. To address these challenges, we propose an efficient lightweight attention network (ELANet) for the skin lesion segmentation task. In ELANet, two different attention mechanisms of the bilateral residual module (BRM) can achieve complementary information, which enhances the sensitivity to features in spatial and channel dimensions, respectively, and then multiple BRMs are stacked for efficient feature extraction of the input information. In addition, the network acquires global information and improves segmentation accuracy by putting feature maps of different scales through multi-scale attention fusion (MAF) operations. Finally, we evaluate the performance of ELANet on three publicly available datasets, ISIC2016, ISIC2017, and ISIC2018, and the experimental results show that our algorithm can achieve 89.87%, 81.85%, and 82.87% of the mIoU on the three datasets with a parametric of 0.459 M, which is an excellent balance between accuracy and lightness and is superior to many existing segmentation methods.

Design and optimization of microchannel for enhancement of the intensity of induced signal in particle/cell impedance measurement.

The measured impedance signal of a single particle or cell in a microchannel is of the μA level, which is a challenge for measuring such weak signals. Therefore, it is necessary to improve the intensity for expanding the applications of impedance measurement. In this paper, we analyzed the impact of geometric parameters of microchannel on output signal intensity by using the three-dimensional finite element method. In comparison to conventional microchannels, which are distributed at a uniform height, the microchannels in this design use the height difference to enhance the signal intensity. By analyzing the effects of the geometric dimensions of the constriction channel, main channel height, radius of particles, types of cells, shapes of particles with different ellipticities, and particles spacing on the current signal, we concluded the optimal dimensions of these parameters to improve the intensity of the induced current signal. Through the fabrication of the optimized size of device and experimental demonstration, it is verified that the current signal intensity caused by the particle with a diameter of 10 µm is nearly twice that of the conventional structure with a height of 20 µm, which proves the correctness of the optimization results and the feasibility of this work. In addition, the performance of the device was verified by measuring the mixtures of different size particles as well as non-viable and viable yeast cells.

Short-term wind power forecasting based on multi-scale receptive field-mixer and conditional mixture copula

Integrating renewable wind power (WP) into the grid exacerbates variability and challenges reliability. Establishing an effective forecasting system is crucial for risk avoidance, but achieving effective, interpretable, and accurate predictions remains an obstacle due to the high volatility and uncertainty of WP. This study proposes an innovative and trustworthy wind power forecasting system combined with point and interval predictions. Specifically, the adaptive multi-scale convolutional receptive field (AMSCF) combined with gated recurrent unit (GRU) for WP point prediction is proposed to capture the spatiotemporal dependencies of WP generation. This AMSCF method uses different size receptive fields, fuses them using channel dimension concatenation, operates pooling layer and adaptive fusion through field attention, which extracts hierarchical features for WP forecasting. Then conditional mixture Copula (Con-mCopula) function integrated with multiobjective optimizer is established for WP interval prediction, which reduces the WP interval prediction accuracy error as some information neglected. The wind power cluster dataset from Germany was selected for experiments. Based on the results, for both datasets, the designed prediction system can achieve better point and interval prediction performance resulting in a reduction of mean absolute error (MAE) by approximately 40 % compared to other existing components in the market. This study deconstructs the “Black Box” in deep network for energy forecasting through interpretability analysis, which accelerates the development of AI technology in WP forecasting system and provides decision-makers with trustworthy WP forecasting assistance.

Characterizing Acoustic Behavior of Silicon Microchannels Separated by a Porous Wall.

Lateral flow membrane microdevices are widely used for chromatographic separation processes and diagnostics. The separation performance of microfluidic lateral membrane devices is determined by mass transfer limitations in the membrane, and in the liquid phase, mass transfer resistance is dependent on the channel dimensions and transport properties of the species separated by the membrane. We present a novel approach based on an active bulk acoustic wave (BAW) mixing method to enhance lateral transport in micromachined silicon devices. BAWs have been previously applied in channels for mixing and trapping cells and particles in single channels, but this is, to the best of our knowledge, the first instance of their application in membrane devices. Our findings demonstrate that optimal resonance is achieved with minimal influence of the pore configuration on the average lateral flow. This has practical implications for the design of microfluidic devices, as the channels connected through porous walls under the acoustic streaming act as 760 µm-wide channels rather than two 375 µm-wide channels in the context of matching the standing pressure wave criteria of the piezoelectric transducer. However, the roughness of the microchannel walls does seem to play a significant role in mixing. A roughened (black silicon) wall results in a threefold increase in average streaming flow in BAW mode, suggesting potential avenues for further optimization.

(HTBNet)Arbitrary Shape Scene Text Detection with Binarization of Hyperbolic Tangent and Cross-Entropy.

The existing segmentation-based scene text detection methods mostly need complicated post-processing, and the post-processing operation is separated from the training process, which greatly reduces the detection performance. The previous method, DBNet, successfully simplified post-processing and integrated post-processing into a segmentation network. However, the training process of the model took a long time for 1200 epochs and the sensitivity to texts of various scales was lacking, leading to some text instances being missed. Considering the above two problems, we design the text detection Network with Binarization of Hyperbolic Tangent (HTBNet). First of all, we propose the Binarization of Hyperbolic Tangent (HTB), optimized along with which the segmentation network can expedite the initial convergent speed by reducing the number of epochs from 1200 to 600. Because features of different channels in the same scale feature map focus on the information of different regions in the image, to better represent the important features of all objects in the image, we devise the Multi-Scale Channel Attention (MSCA). Meanwhile, considering that multi-scale objects in the image cannot be simultaneously detected, we propose a novel module named Fused Module with Channel and Spatial (FMCS), which can fuse the multi-scale feature maps from channel and spatial dimensions. Finally, we adopt cross-entropy as the loss function, which measures the difference between predicted values and ground truths. The experimental results show that HTBNet, compared with lightweight models, has achieved competitive performance and speed on Total-Text (F-measure:86.0%, FPS:30) and MSRA-TD500 (F-measure:87.5%, FPS:30).

Forest fire detection utilizing ghost Swin transformer with attention and auxiliary geometric loss

Forest fires are a devastating natural disaster. Existing fire detection models face limitations in dataset availability, multi-scale feature extraction, and locating obscured or small flames and smoke. To address these issues, we develop a dataset containing real and synthetic forest fire images, sourced from a UAV (Unmanned Aerial Vehicle) perspective. Additionally, we propose the Ghost Convolution Swin Transformer (GCST) module to extract multi-scale flame and smoke features from different receptive fields by integrating parallel Ghost convolution and Swin Transformer. Subsequently, we design a lightweight reparameterized rotation attention module, which captures interactions across channel and spatial dimensions to suppress background noise and focus on obscured flames and smoke. Finally, we introduce a loss function, called Efficient Auxiliary Geometric Intersection over Union (EAGIoU), which employs an auxiliary bounding box to accelerate the model's convergence while integrating the geometrical principles of the predicted and real bounding boxes to accurately locate small flames and smoke. Extensive experimental results demonstrate that our method achieves 75.2 % mAP@0.5 and 42 % mAP@0.5:0.95 with a frame rate of 239 frames per second, indicating a significant improvement in accuracy and real-time performance compared to state-of-the-art techniques. The code and datasets are available at https://github.com/luckylil/forest-fire-detection.

Avionics Module Fault Diagnosis Algorithm Based on Hybrid Attention Adaptive Multi-Scale Temporal Convolution Network.

Since the reliability of the avionics module is crucial for aircraft safety, the fault diagnosis and health management of this module are particularly significant. While deep learning-based prognostics and health management (PHM) methods exhibit highly accurate fault diagnosis, they have disadvantages such as inefficient data feature extraction and insufficient generalization capability, as well as a lack of avionics module fault data. Consequently, this study first employs fault injection to simulate various fault types of the avionics module and performs data enhancement to construct the P2020 communications processor fault dataset. Subsequently, a multichannel fault diagnosis method, the Hybrid Attention Adaptive Multi-scale Temporal Convolution Network (HAAMTCN) for the integrated functional circuit module of the avionics module, is proposed, which adaptively constructs the optimal size of the convolutional kernel to efficiently extract features of avionics module fault signals with large information entropy. Further, the combined use of the Interaction Channel Attention (ICA) module and the Hierarchical Block Temporal Attention (HBTA) module results in the HAAMTCN to pay more attention to the critical information in the channel dimension and time step dimension. The experimental results show that the HAAMTCN achieves an accuracy of 99.64% in the avionics module fault classification task which proves our method achieves better performance in comparison with existing methods.

Machining quality prediction of multi-feature parts using integrated multi-source domain dynamic adaptive transfer learning

Machining quality prediction of multi-feature parts has been a challenging problem because of small dataset and inconsistent quality data distribution with respect to each machining feature. Transfer learning that leverages knowledge of one task and can be repurposed on another task seems a good solution for this purpose. However, traditional transfer learning typically has a single source domain and a target domain, which limits its applications in multi-source scenarios (e.g., multi-feature). To solve this issue, this paper proposes a novel integrated multi-source domain dynamic adaptive transfer learning (IMD-DATL) framework for machining quality prediction of multi-feature part machining systems. Specifically, a domain-sample similarity double matching multi-source domain integration method is designed to construct the integration knowledge transfer from multiple source domains to the target domain. A residual feature extraction network based on sample entropy-dynamic channel double-layer attention structure and a fine-grained transferable feature attention module are designed. These three attentions are used to improve the feature learning ability and the adaptation level to the predicted object in the three dimensions of sample, channel and data feature. Finally, multiple sets of comparative experiments in thin-walled part machining systems confirm the effectiveness and superiority of the proposed method for cross-domain quality prediction. Compared with other traditional transfer learning methods, the MAE, RMSE and Score on average of this method are increased by 5.47 %, 4.59 % and 4.84 %, respectively, compared with other multi-source domain adaptation methods, the MAE, RMSE and Score on average of this method are increased by 7.13 %, 7.37 % and 6.52 %, respectively.

GAN-Generated Face Detection Based on Multiple Attention Mechanism and Relational Embedding

The rapid development of the Generative Adversarial Network (GAN) makes generated face images more and more visually indistinguishable, and the detection performance of previous methods will degrade seriously when the testing samples are out-of-sample datasets or have been post-processed. To address the above problems, we propose a new relational embedding network based on “what to observe” and “where to attend” from a relational perspective for the task of generated face detection. In addition, we designed two attention modules to effectively utilize global and local features. Specifically, the dual-self attention module selectively enhances the representation of local features through both image space and channel dimensions. The cross-correlation attention module computes similarity between images to capture the global information of the output in the image. We conducted extensive experiments to validate our method, and the proposed algorithm can effectively extract the correlations between features and achieve satisfactory generalization and robustness in generating face detection. In addition, we also explored the design of the model structure and the inspection performance on more categories of generated images (not limited to faces). The results show that RENet also has good detection performance on datasets other than faces.

A hybrid deep learning model towards fault diagnosis of drilling pump

This paper proposes a novel method namely WaveletKernelNet-Convolutional Block Attention Module-BiLSTM for intelligent fault diagnosis of drilling pumps. Initially, the random forest method is applied to determine the target signals that can reflect the fault characteristics of drilling pumps. Accordingly, the WaveletKernelNet-Convolutional Block Attention Module Net is constructed for noise reduction and fault feature extraction based on signals. The Convolutional Block Attention Module embedded in WaveletKernelNet-CBAM adjusts the weight and enhances the feature representation of channel and spatial dimension. Finally, the Bidirectional Long-Short Term Memory concept is introduced to enhance the ability of the model to process time series data. Upon constructing the network, a Bayesian optimization algorithm is utilized to ascertain and fine-tune the ideal hyperparameters, thereby ensuring the network reaches its optimal performance level. With the hybrid deep learning model presented, an accurate fault diagnosis of a real five-cylinder drilling pump is carried out and the results confirmed its applicability and reliability. Two sets of comparative experiments validated the superiority of the proposed method. Additionally, the generalizability of the model is verified through domain adaptation experiments. The proposed method contributes to the safe production of the oil and gas sector by providing accurate and robust fault diagnosis of industrial equipment.

Multi-target domain adaptation intelligent diagnosis method for rotating machinery based on multi-source attention mechanism and mixup feature augmentation

Intelligent diagnostic methods for identifying faults in rotating machinery, based on domain adaptation, have garnered significant attention. However, most current domain adaptation approaches are primarily designed for single-source domain and single-target domain (SSST) applications. There is a dearth of domain adaptation approaches tailored for single-source to multi-target domains (SSMT). In contrast to SSST, SSMT takes a more comprehensive approach by considering relationships across multiple target domains. This approach offers increased versatility and a broader range of potential applications. To address this, an end-to-end multi-target adversarial subdomain adaptation method is proposed that leverages attention mechanism data fusion and mixup feature augmentation. Firstly, the attention mechanism is used to fuse data from different sensors in both channel and spatial dimensions. Subsequently, a mixup-based feature augmentation method is proposed for multi-target domain adaptation. The method is combined with subdomain adaptation and domain discrimination to further reduce the distributional differences between the source and various target domains while relieving the overfitting problem during domain adaptation. Finally, with the above approach, a robust and stable model for multiple target domain fault diagnosis can be trained. Our experimental results illustrate that our approach has higher accuracy and robustness compared to several popular domain adaptation methods.

Elasto-inertial focusing and particle migration in high aspect ratio microchannels for high-throughput separation

The combination of flow elasticity and inertia has emerged as a viable tool for focusing and manipulating particles using microfluidics. Although there is considerable interest in the field of elasto-inertial microfluidics owing to its potential applications, research on particle focusing has been mostly limited to low Reynolds numbers (Re<1), and particle migration toward equilibrium positions has not been extensively examined. In this work, we thoroughly studied particle focusing on the dynamic range of flow rates and particle migration using straight microchannels with a single inlet high aspect ratio. We initially explored several parameters that had an impact on particle focusing, such as the particle size, channel dimensions, concentration of viscoelastic fluid, and flow rate. Our experimental work covered a wide range of dimensionless numbers (0.05 < Reynolds number < 85, 1.5 < Weissenberg number < 3800, 5 < Elasticity number < 470) using 3, 5, 7, and 10 µm particles. Our results showed that the particle size played a dominant role, and by tuning the parameters, particle focusing could be achieved at Reynolds numbers ranging from 0.2 (1 µL/min) to 85 (250 µL/min). Furthermore, we numerically and experimentally studied particle migration and reported differential particle migration for high-resolution separations of 5 µm, 7 µm and 10 µm particles in a sheathless flow at a throughput of 150 µL/min. Our work elucidates the complex particle transport in elasto-inertial flows and has great potential for the development of high-throughput and high-resolution particle separation for biomedical and environmental applications.

Object Detection in Multispectral Remote Sensing Images Based on Cross-Modal Cross-Attention.

In complex environments a single visible image is not good enough to perceive the environment, this paper proposes a novel dual-stream real-time detector designed for target detection in extreme environments such as nighttime and fog, which is able to efficiently utilise both visible and infrared images to achieve Fast All-Weatherenvironment sensing (FAWDet). Firstly, in order to allow the network to process information from different modalities simultaneously, this paper expands the state-of-the-art end-to-end detector YOLOv8, the backbone is expanded in parallel as a dual stream. Then, for purpose of avoid information loss in the process of network deepening, a cross-modal feature enhancement module is designed in this study, which enhances each modal feature by cross-modal attention mechanisms, thus effectively avoiding information loss and improving the detection capability of small targets. In addition, for the significant differences between modal features, this paper proposes a three-stage fusion strategy to optimise the feature integration through the fusion of spatial, channel and overall dimensions. It is worth mentioning that the cross-modal feature fusion module adopts an end-to-end training approach. Extensive experiments on two datasets validate that the proposed method achieves state-of-the-art performance in detecting small targets. The cross-modal real-time detector in this study not only demonstrates excellent stability and robust detection performance, but also provides a new solution for target detection techniques in extreme environments.

An optimizing study of silicon-based microchannels for enhanced thermal transfer

The thermal transport efficiency of microchannel heat sinks is critically influenced by the solid-liquid interface thermal resistance, an aspect that has historically been underappreciated in the design of these systems. We introduce a hybrid model that combines molecular dynamics simulations with computational fluid dynamics to rigorously investigate the interfacial heat transfer characteristics of microchannels. Our methodology involves constructing a nanochannel composed of silicon atoms and water molecules to accurately measure the thermal resistance at the Si–H2O interface. We then extend our analysis by developing two-dimensional rectangular microchannel models. These models are utilized to examine the effects of thermal resistance, channel dimensions, and inlet velocity on the overall interfacial heat transfer rate. Our findings indicate a pronounced size dependency in the variation of the total heat transfer rate, highlighting the critical role of scale in these systems. Significantly, enhancements in heat transfer efficiency are observed with an increase in inlet velocity and a reduction in thermal resistance. These results underscore the importance of optimizing both physical and operational parameters to enhance the efficiency of microchannel heat sinks. The findings are expected to inform future designs and promote more efficient thermal management solutions in various applications.

Multi-scale Segformer: image segmentation algorithm for electric vehicles in fire accident

ABSTRACT Electric vehicles (EVs) have played a significant role in sustainability, and EVs fire accidents have raised doubts in recent years. To solve the mobile analytical equipment limitation in EVs fire accident and help staff receive prompt results at on-spot inspection, we provide a lightweight but accurate Transformer that can ideally adapt to the mobile environment. First, we built on the simple Segformer and extended it to aggregate the representations of amorphous objects, such as fire traces, in image recognition. Second, we used shunt-based self-attention (SSA) to enhance the model for capturing multi-scale contextual information and help distinguish the deformed level of EVs after combustion. Third, we redesigned a simple multi-level information aggregation (MIA) decoder to obtain the relationship between pixels in the channel dimensions by a weighted aggregation. Furthermore, to foster image trace recognition, we put forwards and evaluated the accuracy of models on electric vehicle fire traces (EVFTrace), a dataset of images of burnt EVs. On EVFTrace, the mean intersection over union (mIoU) achieves 72.24%. The float point operations (Flops) and parameters (Params) achieve 114.83 G and 89.5 M. Our model shows excellent efficiency and accuracy for burnt EVs segmentation tasks.

Nested attention network based on category contexts learning for semantic segmentation

The attention mechanism is widely used in the field of semantic segmentation, due to the fact that it can be used to obtain effective long-distance dependencies by assigning different weights to objects according to different tasks. We propose a novel Nested Attention Network (NANet) for semantic segmentation, which combines Feature Category Attention (FCA) and Channel Relationship Attention (CRA) to effectively aggregate same-category contexts in both spatial and channel dimensions. Specifically, FCA captures the dependencies between spatial pixel features and categories to achieve the aggregation of features of the same category. CRA further captures the channel relationships on the output of FCA to obtain richer contexts. Numerous experiments have shown that NANet has a lower number of parameters and computational complexity than other state-of-the-art methods, and is a lightweight model with a lower total number of floating-point operations. We evaluated the performance of NANet on three datasets: Cityscapes, PASCAL VOC 2012, and ADE20K, and the experimental results show that NANet obtains promising results, reaching a performance of 82.6% on the Cityscapes test set.

MA‐ResUNet: Multi‐attention optic cup and optic disc segmentation based on improved U‐Net

AbstractGlaucoma poses a significant threat to vision, capable of causing irreversible damage and, in severe instances, leading to permanent blindness. Accurate optic cup (OC) and optic disc (OD) segmentation are essential in glaucoma screening. In this study, a novel OC and OD segmentation approach is proposed. Based on U‐Net, it is optimized by introducing cardinality dimensions. Moreover, attention gates are implemented to reinforce salient features while suppressing irrelevant information. Additionally, a convolutional block attention module (CBAM) is integrated into the decoder segment. This fusion hones in on effective information in both channel and spatial dimensions. Meanwhile, an image processing procedure is proposed for image normalization and enhancement. All of these increase the accuracy of the model. This model is evaluated on the ORIGA and REFUGE datasets, demonstrating the model's superiority in OC and OD segmentation compared to the state‐of‐the‐art methods. Additionally, after the proposed image processing, cup‐to‐disc ratio (CDR) prediction on a batch of 155 in‐house fundus images yields an absolute CDR error of 0.099, which is reduced by 0.04 compared to the case where only conventional processing was performed.

EPSViTs: A hybrid architecture for image classification based on parameter-shared multi-head self-attention

Vision transformers have been successfully applied to image recognition tasks due to their ability to capture long-range dependencies within an image. However, they still suffer from weak local feature extraction, easy loss of channel interaction information in one-dimensional multi-head self-attention modeling, and large number of parameters. This paper proposes a lightweight image classification hybrid architecture named EPSViTs (Efficient Parameter Shared Transformer, EPSViTs). Firstly, a new local feature extraction module is designed to effectively enhance the expression of local features. Secondly, using the parameter sharing approach, a lightweight multi-head self-attention module based on information interaction is designed, which can globally model the image from both spatial and channel dimensions, and mine the potential correlation of the image in space and channel. Extensive experiments are conducted on three public datasets, a subset of ImageNet, Cifar100 and APTOS2019, a private dataset Mushroom66, and the results show that the hybrid architecture EPSViTs proposed in this paper based on parameter sharing for multi-head self-attentive image classification has obvious advantages, especially on a subset of ImageNet to reach 89.18%, which is a 3.8% improvement compared to Edgevits_xxs, verifying the effectiveness of the model.

Channel-Enhanced Contrastive Cross-Domain Sequential Recommendation

Sequential recommendation help users find interesting items by modeling the dynamic user-item interaction sequences. Due to the data sparseness problem, cross-domain sequential recommendation (CDSR) are proposed. CDSR explore rich data from a source domain to improve performance of the target domain. However, most of the existing CDSR methods are difficult to capture the temporal context of sequences and only learn user preference based on interactions of single domain, which leads to suboptimal performance. To address these shortcomings, we propose a channel-enhanced contrastive cross-domain sequential recommendation model (C3DSR). To be specific, (1) we design a feature extractor, which extends attention to the channel dimension, to extract the user’s channel feature and capture the temporal contextual relationships between sequences. Then we calculate the weights of each channel by using three SE-Res2Blocks and multiply it with the channel feature to obtain user preference. (2) We concatenate the user’s single-domain representation, the cross-domain representation, and the user features to make CDSR. Contrastive learning is leveraged to enhance mutual information between two domains. Experimental results show that the proposed model achieves the significant improvement of performance compared with other CDSR models on Amazon and HVIDEO datasets.

Comparison of the induced flow obtained with 2D and 3D CFD simulations of a solar chimney at different width-to-gap ratios

Solar chimneys are among the most common methods for natural ventilation of buildings. Computational Fluid Dynamics (CFD) has been widely applied in research and design of solar chimneys. Most of the previous studies are based on 2D CFD models which ignore the effects of the width (the third dimension) of the air channel. In addition, the applicability of 2D models for the solar chimneys with a low width-to-gap ratio is still questionable. This study investigates both 2D and 3D CFD models for window-sized vertical solar chimneys. The 2D model is applied to the domain comprising the central plane of the channel gap ( G) and height ( H) while the 3D model is applied to the domain consisting of the channel gap, height, and width ( W). The flow fields, flow rates and thermal efficiencies computed with the 2D and 3D models at different heights, gaps, and widths are compared. The results show that W/ G is the most crucial factor. The effects of the side walls are significant at a low W/ G but gradually diminishes as W/ G increases. At W/ G = 15, the side wall effects are confined to a region of about 2.6% W. Particularly, for W/ G&gt;8, the differences between the 2D and 3D flow rates and thermal efficiencies are within ±5%. These findings offer a reference for researchers and engineers to select between 2D and 3D CFD models for a specific solar chimney.