Search

The challenges in citywide traffic flow are intricate, encompassing various factors like temporal and spatial dependencies, holidays, and weather. Despite the complexity, there are still research gaps in effectively incorporating these spatio‐temporal relations through deep learning. Addressing these gaps is crucial for tackling issues such as traffic congestion, public safety, and efficient traffic management within cities. This paper underscores notable research gaps, including the development of models capable of handling both local and global traffic flow patterns, integrating multi‐modal data sources, and effectively managing spatio‐temporal dependencies. In this paper, we proposed a novel model named 3D spatial–temporal‐based adaptive modeling graph convolutional network (3D(STAMGCN)) that addresses for traffic flow data in better periodicity modeling. In contrast to earlier studies, 3D(STAMGCN) approaches the task of traffic flow prediction as a periodic residual learning problem. This is achieved by capturing the input variation between historical time segments and the anticipated output for future time segments. Forecasting traffic flow, as opposed to a direct approach, is significantly simpler when focusing on learning more stationary deviations. This, in turn, aids in the training of the model. Nevertheless, the networks enable residual generation at each time interval through learned variations between future conditions and their corresponding weekly observations. Consequently, this significantly contributes to achieving more accurate forecasts for multiple steps ahead. We executed extensive experiments on two real‐world datasets and compared the performance of our model to state‐of‐the‐art (SOTA) techniques. [ABSTRACT FROM AUTHOR]

Objective. Given the different noise distribution information of global and local magnetic resonance (MR) images, this study aims to extend the current work on convolutional neural networks that preserve global structure and local details in MR image denoising tasks. Approach. This study proposed a parallel and serial network for denoising 3D MR images, called 3D-PSNet. We use the residual depthwise separable convolution block to learn the local information of the feature map, reduce the network parameters, and thus improve the training speed and parameter efficiency. In addition, we consider the feature extraction of the global image and utilize residual dilated convolution to process the feature map to expand the receptive field of the network and avoid the loss of global information. Finally, we combine both of them to form a parallel network. What's more, we integrate reinforced residual convolution blocks with dense connections to form serial network branches, which can remove redundant information and refine features to further obtain accurate noise information. Main results. The peak signal-to-noise ratio, structural similarity index measure, and root mean square error metrics of 3D-PSNet are as high as 47.79%, 99.81%, and 0.40%, respectively, achieving competitive denoising effect on three public datasets. The ablation experiments demonstrated the effectiveness of all the designed modules regarding all the evaluated metrics in both datasets. Significance. The proposed 3D-PSNet takes advantage of multi-scale receptive fields, local feature extraction and residual dense connections to more effectively restore the global structure and local fine features in MR images, and is expected to help doctors quickly and accurately diagnose patients' conditions. [ABSTRACT FROM AUTHOR]

Seismic data with a high signal-to-noise ratio is beneficial in the inversion and interpretation. Thus, denoising is an indispensable step in the seismic data processing. Traditional denoising methods based on prior knowledge are susceptible to the influence of the hypothesis model and parameters. In contrast, deep learning-based denoising methods can extract deep features from the data autonomously and generate a sophisticated denoising model through adaptive learning. However, these methods generally learn a specific model for each noise level, which results in poor representation ability and suboptimal denoising efficacy when applied to seismic data with different noise levels. To address this issue, we propose a denoising method based on a two-stage convolutional neural network (TSCNN). The TSCNN comprises an estimation subnet (ES) and a denoising subnet (DS). The ES employs a multilayer CNN to estimate noise levels, and the DS performs noise suppression on noisy seismic data based on the ES estimation of the noise distribution. In addition, attention mechanisms are implemented in the proposed network to efficiently extract noise information hidden in complex backgrounds. The TSCNN also adopts the L1 loss function to enhance the generalization ability and denoising outcome of the model, and a residual learning scheme is utilized to solve the problem of network degradations. Experimental results demonstrate that the proposed method can preserve event features more accurately and outperforms existing methods in terms of signal-to-noise ratio and generalization ability. [ABSTRACT FROM AUTHOR]

Copyright of Chinese Journal of Liquid Crystal & Displays is the property of Chinese Journal of Liquid Crystal & Displays and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The environment and agricultural output are being challenged by weeds and crops. Innovative advancements in alternative weed control techniques that seek to reduce the need on herbicides have been spurred by the growing demand for sustainable weed control methods. Having weed recognition that is sufficiently successful is a hurdle to the implementation of these methods for selective in-crop application. Deep learning has shown remarkable promise in a variety of vision tasks, leading to the development of several effective image-based weeds and crops identification systems. This study looks at the newest developments in deep learning techniques for pixel-wise semantic segmentation of identifying crops and weeds. The hardest problem is semantic segmentation-based weeds and crops recognition, which needs to be solved for smart farming to work well. The goal is to give each pixel of a picture its own class name. Deep learning for smart farming has a number of useful applications, one of the most important of which is identifying the exact position of crops and weeds on farms. There are already quite complex systems for separating weeds and crops, with millions of factors that need more time to train. To get around these problems, we suggest an AgriResUpNet design based on deep learning. It is a carefully put together mix of the U-Net and residual learning frameworks. To see how well the suggested model works on the crop and weed GitHub dataset in terms of pixel accuracy, precision, f1-sore, and IoU measures. The suggested network is tested and compared with other cutting-edge networks using a crop and weed dataset that is open to the public. Also, the AgriResUpNet model gave an IoU of 97.58% for weeds and 94.82% for crops. In addition, the data show that the AgriResUpNet can easily find both crops and weeds. This suggests that this design is better for finding weeds early in the growing season. It was shown in experiments and comparisons that the suggested network does better than existing designs in intersection over union (IoU) and F1-score. [ABSTRACT FROM AUTHOR]

Surface wave suppression plays a vital role in enhancing the quality of reflection seismic exploration. Convolutional neural networks (CNNs) can adaptively learn the characteristics of effective signals and surface waves. However, CNN has limited receptive fields and cannot effectively reuse features. When surface waves and effective signals overlap heavily, CNN struggles to preserve effective signals effectively. In this paper, we propose a deformable convolutional wavelet transform network (DCWTN) with residual dense blocks to suppress surface waves. DCWTN contains three types of modules: (1) The deformable convolution module (DCM) is designed to expand the receptive field and enhance seismic events continuity. (2) The wavelet transform enhancement module (WTEM) combines a wavelet transform and a residual dense block to suppress surface waves. It performs multi-scale feature extraction on surface waves according to their time–frequency characteristics to recover detailed information on overlapping parts. (3) The residual dense convolution module (RDCM) is designed for feature enhancement and further refinement of the acquired features. Experimental results show that DCWTN retains more effective signals than four popular methods. [ABSTRACT FROM AUTHOR]

Effective image denoising is vital for accurate biomedical image analysis, serving as a foundational step for subsequent diagnostic and analytical tasks. Medical diagnostic instruments often introduce noise during image acquisition, necessitating robust denoising algorithms to ensure diagnostic accuracy. Although deep learning has shown potential in this area, existing solutions often use resized natural images for training deep learning models. However, maintaining picture details is crucial for accurate analysis because medical images are usually high-resolution and contain significant pixel-level features. To address this challenge, a novel approach of preparing training data by dividing images into smaller square tiles is proposed in this study. This study introduces a deep residual learning-based model, utilizing multiple residual learning blocks and rectified linear unit activations, to perform end-to-end mapping from noisy inputs to denoised outputs. The proposed method was evaluated on 50 degraded images, showing superior performance with an Structural Similarity Index (SSIM) of 0.9844 for mammogram images, surpassing the DCNN approach, which averaged an SSIM of 0.9811. Our model achieves enhanced peak signal-to-noise ratio (PSNR) and SSIM in X-ray image denoising, underscoring the effectiveness of training image partitioning over traditional resizing methods. This work highlights the importance of preserving high-resolution details in medical imaging and introduces a novel data preparation technique that significantly improves denoising outcomes. [ABSTRACT FROM AUTHOR]

Trajectory prediction for heterogeneous traffic agents plays a crucial role in ensuring the safety and efficiency of automated driving in highly interactive traffic environments. Numerous studies in this area have focused on physics-based approaches because they can clearly interpret the dynamic evolution of trajectories. However, physics-based methods often suffer from limited accuracy. Recent learning-based methods have demonstrated better performance, but they cannot be fully trusted due to the insufficient incorporation of physical constraints. To mitigate the limitations of purely physics-based and learning-based approaches, this study proposes a kinematics-aware multigraph attention network (KA-MGAT) that incorporates physics models into a deep learning framework to improve the learning process of neural networks. Besides, we propose a residual prediction module to further refine the trajectory predictions and address the limitations arising from simplified assumptions in kinematic models. We evaluate our proposed model through experiments on two challenging trajectory datasets, namely, ApolloScape and NGSIM. Our findings from the experiments demonstrate that our model outperforms various kinematics-agnostic models with respect to prediction accuracy and learning efficiency.

Monitoring air pollution is important for human health and the environment. Previous studies on the prediction of air pollutants from satellite images have employed machine learning, yet there are few enhancements to the constructure of model. Moreover, the existing models have been successful in predicting pollutants like PM2.5, PM10, and O3. They have not been as effective in predicting other primary air pollutants. To improve the overall prediction performance of the existing model, a novel residual learning backpropagation model, abs. as BresNet, has been proposed in this research. This model has revealed the availability to precisely predict the ground-surface concentration of the six primary air pollutants, PM2.5, PM10, O3, NO2, CO, and SO2, based on the satellite imagery of MODIS AOD. Two of the most commonly used machine learning models so far, viz. the multilayer backpropagation neural network (MLBPN) and random forest (RF), were employed as the control. In the conducted experiments, the proposed BresNet model demonstrated significant improvements of 18.75%/31.94%, 33.82%/85.71%, 15.00%/35.29%, 39.06%/134.21%, 23.23%/68.00%, and 137.14%/260.87% in terms of R2 for the six primary air pollutants, compared to the RF/MLBPN model. Moreover, the convergence speed and loss function of the BresNet model compared to that of the MLBPN decreased by 55.15%, revealing superior convergence speed with the lower loss function. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The defect detection in the urban underground utility tunnel faces the challenges of low illuminance and large shadow region. The images collected have problems such as noise and uneven illumination, posing higher requirements for the image feature extraction ability of the network model. A residual feature pyramid attention network based on dense connections (RFPADNet) is proposed in the defect detection of the urban underground comprehensive pipe in this paper. The proposed network consists of three dense blocks, and each dense block uses four residual feature pyramid attention models (RFPAM) as the main feature extractors. Its focus is to utilize residual learning to obtain the fused multi-scale feature information. Soft mask branches are added to the multi-scale channels to enhance the network model's ability to extract positive features. The experimental results show that the network model proposed in this paper has a training accuracy of 97.50%. Compared with the existing network models, it could achieve better results in various test sets, which has good feasibility and practicality. [ABSTRACT FROM AUTHOR]

Chest x-rays (CXRs) are broadly used in clinical practice to diagnose pulmonary diseases. Developing reliable computer-aided diagnosis (CAD) systems to automate the interpretation of CXRs can save medical practitioners time and improve diagnosis accuracy. Besides, segmenting lungs accurately plays a vital role in CAD systems. However, several challenges hamper the generalizability of current automatic lung segmentation approaches. These include the high varied shape and scale of lungs, the device-specific artifacts in CXRs, and the false positive generation. This study aims to address these issues by proposing a new attention-gated residual deep convolutional network based on an enhanced DeepLabV3 + network. The novel sophistications proposed promote the effective learning of intrinsic lung-related features to improve the network robustness against device-specific artifacts and false positive generation. In the encoder, the proposed framework leverages the attention gate mechanism to selectively filter irrelevant lower-level features using high-level multiscale features learned by the Atrous Spatial Pyramid Pooling (ASPP) module. This improves the network adaptability to different scale and shape lungs and reduces its bias to device-specific artifacts. In the decoder, residual convolutions are used to prevent significant information loss during the segmentation map composition. Several models based on the proposed approach are built by employing the EfficientNetB4 and ResNet50 pretrained models as feature extractors and using different public datasets for training and evaluation. Empirical results demonstrated the proposed framework superiority over many state-of-the-art models by achieving 98.03% for the dice coefficient and 96.14% for the Jaccard index without applying any preprocessing or post-processing. [ABSTRACT FROM AUTHOR]

Abstract While deep learning has become the go-to method for image denoising due to its impressive noise removal capabilities, excessive network depth often plagues existing approaches, leading to significant computational burdens. To address this critical bottleneck, we propose a novel lightweight progressive residual and attention mechanism fusion network that effectively alleviates these limitations. This architecture tackles both Gaussian and real-world image noise with exceptional efficacy. Initiated through dense blocks (DB) tasked with discerning the noise distribution, this approach substantially reduces network parameters while comprehensively extracting local image features. The network then adopts a progressive strategy, whereby shallow convolutional features are incrementally integrated with deeper features, establishing a residual fusion framework adept at extracting encompassing global features relevant to noise characteristics. The process concludes by integrating the output feature maps from each DB and the robust edge features from the convolutional attention feature fusion module (CAFFM). These combined elements are then directed to the reconstruction layer, ultimately producing the final denoised image. Empirical analyses conducted in environments characterized by Gaussian white noise and natural noise, spanning noise levels 15–50, indicate a marked enhancement in performance. This assertion is quantitatively corroborated by increased average values in metrics such as Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), and Feature Similarity Index for Color images (FSIMc), outperforming the outcomes of more than 20 existing methods across six varied datasets. Collectively, the network delineated in this research exhibits exceptional adeptness in image denoising. Simultaneously, it adeptly preserves essential image features such as edges and textures, thereby signifying a notable progression in the domain of image processing. The proposed model finds applicability in a range of image-centric domains, encompassing image processing, computer vision, video analysis, and pattern recognition.

Fault detection and diagnosis have always been the key techniques for safe and reliable operation of industrial processes. However, the high dimension and noise of process variables have brought great challenges to the fault detection model. In recent years, due to the powerful feature extraction ability, deep learning has been widely applied in process fault detection and diagnosis. However, these deep neural networks (DNNs) often need a large amount of label data for supervised training, or show poor performance in learning features under unsupervised-learning condition. This paper proposes a new DNN model, a one-dimension residual convolutional auto-encoder (1DRCAE), where unsupervised learning is used to extract representative features from complex industrial processes. 1DRCAE effectively integrates the one-dimensional convolutional kernel with an auto-encoder and is embedded residual learning block for effective feature extraction from one-dimensional data. The two statistics and squared prediction error are generated in the feature space and residual space of 1DRCAE, respectively. Finally, the feasibility and superiority of 1DRCAE are verified on a simulation process, Tennessee Eastman process, Fed-batch fermentation penicillin process, and a real-life case. The convolutional auto-encoder technique provides a new way for feature learning and fault detection on complex industrial processes. [ABSTRACT FROM AUTHOR]

In order to maintain low-rank characteristics, existing low-rank representation methods concentrate on capturing data's low-frequency signals, which are presumed to be the global data structure, and they delete the high ones, which are often a combination of corrupt elements and image edges. Such inefficient preservation of image edges could hamper discriminative details in images, especially in heavy corruptions. This paper proposes a new method, which preserves image edges by finding robust subspace projections from low-rank residuals. It is achieved through a least square minimization of the discrepancy between similar residuals in a manifold learning framework. Edge preserved subspace projections are learned from such residuals by reducing the influence of corrupt ones using a dynamic affinity graph regularization. Furthermore, through our adaptive learning approach, the proposed method can jointly find image intrinsic low-rank representation. Several experimental results in classification and clustering tasks demonstrate the proposed method's effectiveness over state-of-the-art (SOTA) methods. [ABSTRACT FROM AUTHOR]

Defocus blur detection (DBD), a technique for detecting defocus or in-focus pixels in a single image, has been widely used in various fields. Although deep learning-based methods applied to DBD attain superior performance compared to traditional methods that rely on manually-constructed features, these methods cannot distinguish many microscopic details when the images are complex. To address this issue, this work proposes a hybrid CNN-Transformer architecture (TCRL) based on complementary residual learning, which employs global information captured by the Transformer and hierarchical complementary information from the network to optimize DBD. Specifically, to enhance global target detection, our backbone network adopts a CNN-Transformer architecture, where the Transformer effectively drives the network to focus on the global context and thus achieve precise localization. To better detect microscopic details, we combine each convolutional neural network layer with layered complementary information from the network module to optimize the defocus blur detection process. This strategy opposes current schemes that output a binary mask, affording the layered feature-guided learning method to exploit better both low- and high-level information and effectively drive the network to refocus on boundaries and sparse, easily overlooked parts. Additionally, this work also considers the features of the in-focus and defocus pixels within the image. In this complementary model, the information ignored by one side may be learned by the other side, thus enhancing global target detection and local boundary refinement process. The experimental results on three datasets validate the effectiveness and superiority of the developed method. [ABSTRACT FROM AUTHOR]

This study introduces an innovative deep learning model, Residual-EnDecode-Feedforward Attention Mechanism-Long Short-Term Memory (REDF-LSTM), designed to overcome the high uncertainty challenges faced by traditional soil moisture prediction methods. The REDF-LSTM model, by integrating a residual learning encoder–decoder LSTM layer, enhanced LSTM layers, and feedforward attention, not only captures the deep features of time series data but also optimizes the model's ability to identify key influencing factors, including land surface features, atmospheric conditions, and other static environmental variables. Unlike existing methods, the innovation of this model lies in its first-time combination of the residual learning encoder–decoder and feedforward attention mechanisms in the soil moisture prediction field. It delves into the complex patterns of time series through the encoder–decoder structure and accurately locates key influencing factors through the feedforward attention mechanism, significantly improving predictive performance. The choice to combine the feedforward attention mechanism and encoder–decoder with the LSTM model is to fully leverage their advantages in processing complex data sequences and enhancing the model's focus on important features, aiming for more accurate soil moisture prediction. After comparison with current advanced models such as EDLSTM, FAMLSTM, and GANBiLSTM, our REDF-LSTM demonstrated the best performance. Compared to traditional LSTM models, it achieved an average improvement of 13.07% in R2, 20.98% in RMSE, 24.86% in BIAS, and 11.1% in KGE key performance indicators, fully proving its superior predictive capability and potential application value in precision agriculture and ecosystem management. [ABSTRACT FROM AUTHOR]

Video Super Resolution (VSR) applications extensively utilize deep learning-based methods. Several VSR methods primarily focus on improving the fine-patterns within reconstructed video frames. It frequently overlooks the crucial aspect of keeping conformation details, particularly sharpness. Therefore, reconstructed video frames often fail to meet expectations. In this paper, we propose a Conformation Detail-Preserving Network (CDPN) named as SuperVidConform. It focuses on restoring local region features and maintaining the sharper details of video frames. The primary focus of this work is to generate the high-resolution (HR) frame from its corresponding low-resolution (LR). It consists of two parts: (i) The proposed model decomposes confirmation details from the ground-truth HR frames to provide additional information for the super-resolution process, and (ii) These video frames pass to the temporal modelling SR network to learn local region features by residual learning that connects the network intra-frame redundancies within video sequences. The proposed approach is designed and validated using VID4, SPMC, and UDM10 datasets. The experimental results show the proposed model presents an improvement of 0.43 dB (VID4), 0.78 dB (SPMC), and 0.84 dB (UDM10) in terms of PSNR. Further, the CDPN model set new standards for the performance of self-generated surveillance datasets. [ABSTRACT FROM AUTHOR]

OCT (Optical Coherence Tomography) is a non-invasive diagnostic tool for detecting and treating a wide range of retinal diseases. However, the OCT image formation method produces speckle noise, degrading the quality of OCT images significantly, and these low-quality images negatively impact subsequent illness diagnosis. Traditional approaches to remove speckle noise include spatial/transform domain filtering, dictionary learning, or hybridizing these methods. By adopting a hierarchical network topology, deep Convolutional Neural Networks (CNN) have expanded the capacity to harness spatial correlations and extract data at multiple resolutions, making image denoising algorithms more robust. This paper proposes a residual learning-based despeckling CNN architecture (DespNet) for removing speckle noise from OCT images. Trained on 1440 augmented OCT images, DespNet generates the residual images that contain the detailed noise pattern of input images, which, when subtracted from noisy images, results in the denoised version. Quantitative and qualitative analyses have been done, and the experimental results show that the images despeckled using DespNet architecture substantially reduce speckle noise while preserving texture and structure that aids in retinal layer segmentation and consequent illness diagnosis. [ABSTRACT FROM AUTHOR]

Traditional image detail enhancement is local filter-based or global filter-based. In both approaches, the original image is first divided into the base layer and the detail layer, and then the enhanced image is obtained by amplifying the detail layer. Our method is different, and its innovation lies in the special way to get the image detail layer. The detail layer in our method is obtained by updating the residual features, and the updating mechanism is usually based on searching and matching similar patches. However, due to the diversity of image texture features, perfect matching is often not possible. In this paper, the process of searching and matching is treated as a thermodynamic process, where the Metropolis theorem can minimize the internal energy and get the global optimal solution of this task, that is, to find a more suitable feature for a better detail enhancement performance. Extensive experiments have proven that our algorithm can achieve better results in quantitative metrics testing and visual effects evaluation. The source code can be obtained from the website https://github.com/hehesjtu/Metropolis-Theorem. [ABSTRACT FROM AUTHOR]

The precise identification of splice sites is essential for unraveling the structure and function of genes, constituting a pivotal step in the gene annotation process. In this study, we developed a novel deep learning model, DRANetSplicer, that integrates residual learning and attention mechanisms for enhanced accuracy in capturing the intricate features of splice sites. We constructed multiple datasets using the most recent versions of genomic data from three different organisms, Oryza sativa japonica, Arabidopsis thaliana and Homo sapiens. This approach allows us to train models with a richer set of high-quality data. DRANetSplicer outperformed benchmark methods on donor and acceptor splice site datasets, achieving an average accuracy of (96.57%, 95.82%) across the three organisms. Comparative analyses with benchmark methods, including SpliceFinder, Splice2Deep, Deep Splicer, EnsembleSplice, and DNABERT, revealed DRANetSplicer's superior predictive performance, resulting in at least a (4.2%, 11.6%) relative reduction in average error rate. We utilized the DRANetSplicer model trained on O. sativa japonica data to predict splice sites in A. thaliana, achieving accuracies for donor and acceptor sites of (94.89%, 94.25%). These results indicate that DRANetSplicer possesses excellent cross-organism predictive capabilities, with its performance in cross-organism predictions even surpassing that of benchmark methods in non-cross-organism predictions. Cross-organism validation showcased DRANetSplicer's excellence in predicting splice sites across similar organisms, supporting its applicability in gene annotation for understudied organisms. We employed multiple methods to visualize the decision-making process of the model. The visualization results indicate that DRANetSplicer can learn and interpret well-known biological features, further validating its overall performance. Our study systematically examined and confirmed the predictive ability of DRANetSplicer from various levels and perspectives, indicating that its practical application in gene annotation is justified. [ABSTRACT FROM AUTHOR]

Automated nucleus segmentation is considered the gold standard for diagnosing some severe diseases. Accurate instance segmentation of nuclei is still very challenging because of the large number of clustered nuclei, and the different appearance of nuclei for different tissue types. In this paper, a neural network is proposed for fast and accurate instance segmentation of nuclei in histopathology images. The network is inspired by the Unet and residual nets. The main contribution of the proposed model is enhancing the classification accuracy of nuclear boundaries by moderately preserving the spatial features by relatively d the size of feature maps. Then, a proposed 2D convolution layer is used instead of the conventional 3D convolution layer, the core of CNN-based architectures, where the feature maps are first compacted before being convolved by 2D kernel filters. This significantly reduces the processing time and avoids the out of memory problem of the GPU. Also, more features are extracted when getting deeper into the network without degrading the spatial features dramatically. Hence, the number of layers, required to compensate the loss of spatial features, is reduced that also reduces the processing time. The proposed approach is applied to two multi-organ datasets and evaluated by the Aggregated Jaccard Index (AJI), F1-score and the number of frames per second. Also, the formula of AJI is modified to reflect the object- and pixel-level errors more accurately. The proposed model is compared to some state-of-the-art architectures, and it shows better performance in terms of the segmentation speed and accuracy. [ABSTRACT FROM AUTHOR]

Convolutional neural networks (CNNs) are widely popular in the field of image denoising. A large number of CNN-based denoising methods exhibit superior denoising performance in comparison with most conventional denoising schemes. However, some of these approaches extract the noise by stacking many common convolutional layers, which makes them prone to overfitting and causes more loss of image details since the erroneous extraction of non-noise features. A new multi-scale denoising network (MSDNet) is proposed for better tackling these issues, which uses the multi-scale feature information and pixel-wise correlation to effectively remove more noise from noisy images and retain more image details. The denoising effectiveness of MSDNet is specifically attributed to its three key modules, namely multi-scale progressive fusion block (MSPFB), pixel-wise attention block (PWAB) and residual learning (RL), in which MSPFB helps MSDNet capture more useful context information and reduce important information loss caused by ignoring scale inconsistency for capturing more noise from noisy images while maintaining more image details, PWAB facilitates MSDNet to selectively focus on specific image pixels or regions for further effectively capturing noise from noisy images while better preserving image details, and RL is helpful for MSDNet to better address deeper neural network training difficulties and mitigate overfitting. Experimental results demonstrate that MSDNet exhibits superior denoising and single-image deraining performance. [ABSTRACT FROM AUTHOR]

One of the most difficult problems that develop when brain cells start to grow out of control is a brain tumor, which is regarded as the most lethal disease of the century. Finding and identifying malignant brain magnetic resonance imaging (MRI) images is the major challenge before therapy. Researchers have been putting a lot of effort into creating the best method for more accurate real‐world medical image recognition. For manual categorization, it is quite time‐consuming to segment large quantities of MRI data. To mitigate these issues, this paper suggests the information exchange gateway‐based residual UNet (IEGResUNet) model, which uses the ResUNet model as a base model. Additionally, including principal component analysis (PCA) data augmentation will increase the model's efficiency while also enhancing its speed. The IEGResUNet model shows an ablation investigation on three Brats datasets, with and without PCA augmentation. The results demonstrate that IEGResUNet will improve segmentation effectiveness and can also manage the imbalance in input data when PCA data augmentation models are included. The dice score on BraTS 2019 for whole tumor, region of core tumor, and region of enhancing tumor were 0.9083, 0.883, and 0.8106 respectively. Also, on BraTS 2020, the dice score for WT, CT, and ET 0.9083, 0.883, and 0.8106 was respectively. Similarly, on BraTS 2021, the dice score for WT, CT, and ET was 0.8737, 0.8866, and 0.7963 respectively. Comparing against baseline models, the IEGResUNet scored well in terms of dice score and intersection over union. [ABSTRACT FROM AUTHOR]

Existing deep learning-based hyperspectral anomaly detection methods typically perform anomaly detection by reconstructing a clean background. However, for the deep networks, there are many parameters that need to be adjusted. To reduce parameters of network and improve the performance of anomaly detection, a light CNN based on residual learning and background estimation was proposed. Different from traditional methods, the proposed method could directly learn anomaly features rather than background features. First, during the training stage, a background estimation method based on non-central convolution kernels was used to obtain the pseudo-background. Second, to purify the pseudo-background, a pair down-sampling method and a joint loss that combines cross-approximation background loss and consistency loss were proposed. Third, the anomaly matrix was obtained by the difference between the hyperspectral image (HSI) and the pseudo-background. Fourth, a light CNN with three layers was proposed to extract features of the anomaly matrix. Finally, during the prediction stage, anomaly detection results were calculated from the predicted anomaly matrix obtained by light CNN through the Mahalanobis distance. Experiments were conducted with multiple metrics on five real-world datasets. Compared with eight state-of-the-art methods, the proposed method achieved the superior performance in both qualitative and quantitative evaluations.

Electromyographic (EMG) signals have gained popularity for controlling prostheses and exoskeletons, particularly in the field of upper limbs for stroke patients. However, there is a lack of research in the lower limb area, and standardized open-source datasets of lower limb EMG signals, especially recording data of Asian race features, are scarce. Additionally, deep learning algorithms are rarely used for human motion intention recognition based on EMG, especially in the lower limb area. In response to these gaps, we present an open-source benchmark dataset of lower limb EMG with Asian race characteristics and large data volume, the JJ dataset, which includes approximately 13,350 clean EMG segments of 10 gait phases from 15 people. This is the first dataset of its kind to include the nine main muscles of human gait when walking. We used the processed time-domain signal as input and adjusted ResNet-18 as the classification tool. Our research explores and compares multiple key issues in this area, including the comparison of sliding time window method and other preprocessing methods, comparison of time-domain and frequency-domain signal processing effects, cross-subject motion recognition accuracy, and the possibility of using thigh and calf muscles in amputees. Our experiments demonstrate that the adjusted ResNet can achieve significant classification accuracy, with an average accuracy rate of 95.34% for human gait phases. Our research provides a valuable resource for future studies in this area and demonstrates the potential for ResNet as a robust and effective method for lower limb human motion intention pattern recognition.

High dynamic range (HDR) imaging techniques offer photographers the ability to capture the full range of luminance in real-world scenes, overcoming the limitations of capture and display devices. One popular method for creating HDR images is the multiple exposures technique, which involves capturing multiple exposures with regular digital cameras and combining them later to generate an HDR image. In this work, we propose a method called Residual Compressed Exposure Sequences (ResCES) that aims to consolidate all the information from a bracketed sequence into a single JPEG file. Typically, the main image that is to be displayed by a standard image viewer is selected as the middle exposure of the sequence, although any other user-preferred exposure can be selected as well. When needed, the original exposures can be reconstructed from this single JPEG file, enabling their use in a standard HDR workflow. Our proposed approach utilizes a patch-based process, where we store under-exposed, over-exposed, and motion-detected patches while reconstructing other patches through the camera response function to minimize data loss. To further improve the fidelity of the reconstructed exposures, we employ a residual learning model in the last stage of our pipeline, effectively eliminating any artifacts that may occur in its earlier stages. The key innovation of ResCES is its ability to encapsulate the complete set of original exposures within a single JPEG file in an efficient manner, allowing for on-demand reconstruction – a feature that distinguishes it from existing HDR file formats in the literature. The experimental results demonstrate that ResCES achieves a high degree of similarity with respect to the original exposures, as shown by both quantitative and qualitative evaluations. The subjective visual evaluation conducted using 40 participants indicates that ResCES reconstruction results are statistically indistinguishable from the original exposures, while, on average, yielding a 4.5 times storage reduction. This, coupled with the ease of file maintenance, simplifies storing, sharing, and viewing of HDR images.

Structural dynamic response corrupts frequently due to the sensor malfunction. The loss of dynamic response will hinder the structural condition assessment. In recent years, significant efforts have been devoted to recovering the dynamic response during the linear elastic stage of the structure. However, relevant researches on the response recovery of damaged structures are rarely reported due to its strong nonlinearity. With the growing significance of post-disaster structural maintenance, it is critical to develop effective methods for recovering missing data in damaged structures. To this end, this paper proposes a dynamic response recovery method for damaged structures using residual learning enhanced fully convolutional networks (FCN), which can provide a baseline for the recovery of monitoring data in operational civil infrastructure. Specifically, a FCN incorporating residual learning and skip connections is designed to capture high-dimensional nonlinear relationships between input and output channels, thereby achieving the data recovery for any concerned channel. Then, a time–frequency domain evaluation mode is constructed, in which L2 norm is used to measure the difference of recovery results in the time domain, while instantaneous frequency is employed to evaluate the integrity of the spectral information of recovery results. Finally, a destruction test of an experimental arch was conducted, and the acceleration data under different damaged state were collected to investigate the feasibility of the proposed method. Besides, the recovery effects concerning input channel location and quantity, multi-channel response and cross-state recovery are examined. The results show that even in a severely damaged state, the proposed method effectively recovers the missing data. In addition, improving the correlation between input and output channels and increasing the number of input channels can further enhance the recovery accuracy. [ABSTRACT FROM AUTHOR]

The automation of hand hygiene monitoring is critical in healthcare for ensuring clean hands and preventing infectious disease spread. While advancements have been made, existing methods have limitations in accurately detecting and classifying handwashing actions. This paper addresses these limitations and introduces the Residual-Based Multilevel Fused Network (ResMFuse-Net) as a novel approach to automate the quality assurance of hand hygiene procedures. Our model integrates advanced techniques, including feature fusion, model compression, a feature fusion block (FFB), and a modified separable residual block (SE-ResB). The proposed model fused two networks into one trainable feature extraction pipeline, and applies model compression to retain the core blocks that are crucial for propagating strong and robust features while conserving a significant fraction of the computing resources. Additionally, we introduce a FFB that includes ConvLSTM and alpha dropout to learn spatial dependencies, establish correlations between frames in a video, and mitigate overfitting. This paper introduces a SE-ResB, which is a customized residual component composed of separable convolutions and LeakyReLU activation. The SE-ResB is incorporated to handle the fused features and generate a more diverse set of features, leading to considerable performance enhancements. This study also includes an ablation analysis that highlights the importance of each component. The proposed ResMFuse-Net is evaluated on two datasets: a newly created handwashing dataset (451 videos) and a publicly available dataset (656 videos). Achieving a recognition accuracy of 97.61% on the handwashing dataset and 98.69% on the other dataset, the ResMFuse-Net outperforms previous methods with fewer parameters and FLOPs, demonstrating its efficiency and cost-effectiveness. [ABSTRACT FROM AUTHOR]

Deep convolutional neural networks can effectively improve the performance of single-image super-resolution reconstruction. Deep networks tend to achieve better performance than others. However, the deep CNNs will lead to a dramatic increase in the size of parameters, limiting its application on embedding and resource-constrained devices, such as smart phone. To address the common problems of blurred image edges, inflexible convolution kernel size selection and slow convergence during training procedure due to redundant network structure in image super-resolution algorithms, this paper proposes a lightweight single-image super-resolution network that fuses multi-level features. The components are mainly two-level nested residual blocks. To better extract features and reduce the number of parameters, each residual block adopts an asymmetric structure. Firstly, it expands twice and then compresses the number of channels twice. Secondly, in the residual block, the feature information of different channels is weighted and fused by adding an autocorrelation weight unit. The quality of the reconstructed image of the proposed method is superior to the existing image super-resolution reconstruction methods in both subjective perception and objective evaluation indicators, and the reconstruction performance is better when the factor is large. [ABSTRACT FROM AUTHOR]

Lung cancer is one of the leading causes of death worldwide and early detection is crucial to reduce the mortality. A reliable computer-aided diagnosis (CAD) system can help facilitate early detection of malignant nodules. Although existing methods provide adequate classification accuracy, there is still room for further improvement. This study is dedicated to investigating a new CAD scheme for predicting the malignant likelihood of lung nodules in computed tomography (CT) images in light of a deep learning strategy. Conceived from the residual learning and selective kernel, we investigated an efficient residual selective kernel (RSK) block to handle the diversity of lung nodules with various shapes and obscure structures. Founded on this RSK block, we established a multiview RSK network (MRSKNet), to which three anatomical planes in the axial, coronal, and sagittal directions were fed. To reinforce the classification efficiency, seven handcrafted texture features with a filter-like computation strategy were explored, among which the homogeneity (HOM) feature maps are combined with the corresponding intensity CT images for concatenation input, leading to an improved network architecture. Evaluated on the public benchmark Lung Image Database Consortium and Image Database Resource Initiative (LIDC-IDRI) challenge database with ten-fold cross validation of binary classification, our experimental results indicated high area under receiver operating characteristic (AUC) and accuracy scores. A better compromise between recall and specificity was struck using the suggested concatenation strategy comparing to many state-of-the-art approaches. The proposed pulmonary nodule classification framework exhibited great efficacy and achieved a higher AUC of 0.9711. The association of handcrafted texture features with deep learning models is promising in advancing the classification performance. The developed pulmonary nodule CAD network architecture is of potential in facilitating the diagnosis of lung cancer for further image processing applications. [ABSTRACT FROM AUTHOR]

In this paper, we present a denoising method for path tracing using residual learning with convolutional neural networks (CNNs). Noisy artifacts in path tracing are inherited from insufficient sampling, which often generates over- or under-exposed values when integrating the limited bright or dark samples in a pixel. In this paper, we introduce a dual channel residual learning CNNs which separates the over and under-exposed signals in order to provide an efficient denoising filter for the path tracing rendering. Furthermore, we present an advanced CNN comprised of variable-sized kernels in each convolutional layer. Our CNN detects features in different scales providing an adaptive denoising filter capability which is optimal for extracting various contextual details in a complex scene. The experiments demonstrate that our method generates better visual quality than other compared approaches across various rendering effects. [ABSTRACT FROM AUTHOR]

The barren plateau problem in quantum neural networks (QNNs) is a significant challenge that hinders the practical success of QNNs. In this paper, we introduce residual quantum neural networks (ResQNets) as a solution to address this problem. ResQNets are inspired by classical residual neural networks and involve splitting the conventional QNN architecture into multiple quantum nodes, each containing its own parameterized quantum circuit, and introducing residual connections between these nodes. Our study demonstrates the efficacy of ResQNets by comparing their performance with that of conventional QNNs and plain quantum neural networks through multiple training experiments and analyzing the cost function landscapes. Our results show that the incorporation of residual connections results in improved training performance. Therefore, we conclude that ResQNets offer a promising solution to overcome the barren plateau problem in QNNs and provide a potential direction for future research in the field of quantum machine learning. [ABSTRACT FROM AUTHOR]

BACKGROUND: In clinical medicine, low-dose radiographic image noise reduces the quality of the detected image features and may have a negative impact on disease diagnosis. OBJECTIVE: In this study, Adaptive Projection Network (APNet) is proposed to reduce noise from low-dose medical images. METHODS: APNet is developed based on an architecture of the U-shaped network to capture multi-scale data and achieve end-to-end image denoising. To adaptively calibrate important features during information transmission, a residual block of the dual attention method throughout the encoding and decoding phases is integrated. A non-local attention module to separate the noise and texture of the image details by using image adaptive projection during the feature fusion. RESULTS: To verify the effectiveness of APNet, experiments on lung CT images with synthetic noise are performed, and the results demonstrate that the proposed approach outperforms recent methods in both quantitative index and visual quality. In addition, the denoising experiment on the dental CT image is also carried out and it verifies that the network has a certain generalization. CONCLUSIONS: The proposed APNet is an effective method that can reduce image noise and preserve the required image details in low-dose radiographic images. [ABSTRACT FROM AUTHOR]

Images are easily polluted by noise in the process of acquisition and transmission, which will affect people's understanding and utilization of knowledge and information in images. Therefore, image denoising, as a classic problem, has received extensive attention from researchers. At present, many image denoising methods based on deep learning have been proposed and achieved good performance. However, most existing methods are insufficient in acquiring and utilizing crucial information in the image when removing noise under complex image denoising tasks such as blind denoising and real-world denoising, resulting in the loss of fine details in the reconstructed image. To overcome this shortcoming, in this paper, we propose a novel image denoising algorithm combining attention mechanism and residual UNet network, named Att-ResUNet. Specifically, we propose a novel UNet-based image denoising framework, which employs residual enhancement blocks and skip connections to form global–local residuals, which can fuse multi-scale global context and local features to more thoroughly capture and remove hidden noise in the image. A channel attention mechanism is introduced, which can better focus on the crucial information in the image and improve the denoising performance. In addition, we use adaptive average pooling for down-sampling, which can preserve more image structure information, reduce the loss of edge details, and adopt a residual learning strategy to enhance the learning and expressive capabilities of the denoising model. Extensive experiments on several publicly available standard datasets demonstrate the superiority of our method over 15 state-of-the-art methods and achieve excellent denoising performance. Compared with mainstream methods, our method outperforms current state-of-the-art methods by up to 0.76 dB and 1.10 dB on PSNR evaluation metrics on BSD68 and Set12 datasets, respectively. Notably, our method achieves an average PSNR value of 37.88 dB on the CC dataset in real-world denoising experiments, a significant improvement of 2.14 dB over the most advanced methods. [ABSTRACT FROM AUTHOR]

Copyright of Machine Tool & Hydraulics is the property of Guangzhou Mechanical Engineering Research Institute (GMERI) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Computer Engineering & Applications is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Journal of Test & Measurement Technology is the property of Publishing Center of North University of China and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The finite difference forward modeling has been widely used in geophysics exploration and petroleum fields. Because of its high efficiency and easy application for graphical processing units, it has been widely concerned by industry and academia. However, owing to many factors, the problem of numerical dispersion has been an important factor hindering this method. To overcome the numerical dispersion, this paper proposes a method for removing numerical dispersion using deep learning. Unlike the conventional optimized algorithms target to optimize the finite difference coefficients, our strategy is based on big data training to eliminate the dispersion data after seismic data modeling. We design a neural network architecture based on cycle-consistent generative adversarial networks (Cycle-GANs) and residual learning for elastic wave propagation. Under the premise of not significantly increasing the calculation time, we can obtain higher calculation accuracy. Compared with the high-order finite difference algorithm, the calculation time is the advantage of our proposed deep learning method. Tests prove the efficiency and stability of our proposed algorithm. [ABSTRACT FROM AUTHOR]

Sports plays a pivotal role in national development. To accurately predict college students' sports performance and motivate them to improve their physical fitness, this study constructs a sports achievement prediction system by using a U-Net Convolutional Neural Network (CNN) in machine learning. Firstly, the current state of physical education teachers’ instructional proficiency is investigated and analyzed to identify existing problems. Secondly, an improved U-Net-based sports achievement prediction system is proposed. This method enhances the utilization and propagation of network features by incorporating dense connections, thus addressing gradient disappearance issues. Simultaneously, an improved mixed loss function is introduced to alleviate class imbalance. Finally, the effectiveness of the proposed system is validated through testing, demonstrating that the improved U-Net CNN algorithm yields superior results. Specifically, the prediction accuracy of the improved network for sports performance surpasses that of the original U-Net by 4.22 % and exceeds that of DUNet by 5.22 %. Compared with other existing prediction networks, the improved U-Net CNN model exhibits a superior achievement prediction ability. Consequently, the proposed system enhances teaching and learning efficiency and offers insights into applying artificial intelligence technology to smart classroom development.

Medical workers can assess disease progression and create expedient treatment plans with the help of automated and accurate 3Dsegmentation of medical images. DCNNs (Deep convolution neural networks) have been widely used in this work, but their accuracy still needs to be increased, mostly due to their insufficient understanding of 3D environments. This study proposed three dimensional residual networks, ResUNet++, for precise segmentations of three-dimensional medical images where encoders, segmentation decoders, and context residual decoders are used. Two decoders are connected at scale utilizing context attention maps and context residual, the former explicitly learns inter-slice context data and the latter utilizes contexts as attention to increase segmentation accuracy. This model was assessed by using MICCAI 2018 BraTS dataset and, the Pancreas-CT dataset. The BrasTS and Pancreas-CT dataset scales were compared in terms of ET, WT, TC. Moreover, the proposed model was compared with/without boundary loss and validation dice score. The outcomes not only show how effective the suggested 3D residual learning approach is, but also show that the suggested ResUNet++ offers better accuracy compared to six of the top-ranking techniques used for segmenting tumors in the brain.