Adaptive Block Research Articles

Atrial fibrillation medical image encryption algorithm based on deep learning and adaptive block

Atrial fibrillation medical image encryption algorithm based on deep learning and adaptive block

Atrial fibrillation medical image encryption algorithm based on deep learning and adaptive block

Atrial fibrillation medical image encryption algorithm based on deep learning and adaptive block

A Robust Frequency-Domain-Based Graph Adaptive Network for Parkinson's Disease Detection From Gait Data

Parkinson's disease (PD) is a neurodegenerative disease with a high incidence rate. Effective early diagnosis of PD is critical to prevent further deterioration of a patient's condition, where gait abnormalities are important factors for doctors to diagnose PD. Deep learning (DL)-based methods for PD detection using gait information recorded by non-invasive sensors have emerged to assist doctors in accurate and efficient disease diagnosis. However, most existing DL-based PD detection models neglect information in the frequency domain and do not adaptively model the correlation of signals among sensors. Moreover, different people have different gait patterns. Therefore, the generalization capabilities of PD detection models on diversities of individuals' gaits are essential. This work proposes a novel robust frequency-domain-based graph adaptive network (RFdGAD) for PD detection from gait information (i.e., vertical ground reaction force signals recorded by foot sensors). Specifically, the RFdGAD first learns the frequency-domain features of signals from each foot sensor by a frequency representation learning block. Then, the RFdGAD utilizes a graph adaptive network block taking frequency-domain features as input to adaptively learn and exploit the interconnection between different sensor signals for accurate PD detection. Moreover, the RFdGAD is trained by minimizing the proposed Jensen-Shannon divergence-based localized generalization error to improve the generalization performance of RFdGAD on unseen subjects. Experimental results show that the RFdGAD outperforms existing DL-based models for PD detection on three widely used datasets in terms of three metrics, including accuracy, F1-score, and geometric mean.

Driver distraction detection based on lightweight networks and tiny object detection.

Real-time and efficient driver distraction detection is of great importance for road traffic safety and assisted driving. The design of a real-time lightweight model is crucial for in-vehicle edge devices that have limited computational resources. However, most existing approaches focus on lighter and more efficient architectures, ignoring the cost of losing tiny target detection performance that comes with lightweighting. In this paper, we present MTNet, a lightweight detector for driver distraction detection scenarios. MTNet consists of a multidimensional adaptive feature extraction block, a lightweight feature fusion block and utilizes the IoU-NWD weighted loss function, all while considering the accuracy gain of tiny target detection. In the feature extraction component, a lightweight backbone network is employed in conjunction with four attention mechanisms strategically integrated across the kernel space. This approach enhances the performance limits of the lightweight network. The lightweight feature fusion module is designed to reduce computational complexity and memory access. The interaction of channel information is improved through the use of lightweight arithmetic techniques. Additionally, CFSM module and EPIEM module are employed to minimize redundant feature map computations and strike a better balance between model weights and accuracy. Finally, the IoU-NWD weighted loss function is formulated to enable more effective detection of tiny targets. We assess the performance of the proposed method on the LDDB benchmark. The experimental results demonstrate that our proposed method outperforms multiple advanced detection models.

AWFLN: An Adaptive Weighted Feature Learning Network for Pansharpening

Deep learning (DL)-based pansharpening methods have shown great advantages in extracting spectral-spatial features from multispectral (MS) and panchromatic (PAN) images compared with traditional methods. However, most DL-based methods ignore the local inner connection between the source images and high-resolution MS (HRMS) image, which cannot fully extract spectral-spatial information, and attempt to improve the quality of fusion by increasing the complexity of the network. To solve these problems, a lightweight network based on adaptive weighted feature learning (AWFLN) is proposed for pansharpening. Specifically, a novel detail extraction model is first built by exploring the local relationship between HRMS and source images, thereby improving the accuracy of details and the interpretability of the network. Guided by this model, we then design a residual multiple receptive-field structure to fully extract spectral-spatial features of source images. In this structure, an adaptive feature learning block based on spectral-spatial interleaving attention is proposed to adaptively learn the weights of features and improve the accuracy of the extracted details. Finally, the pansharpened result is obtained by a detail injection model in AWFLN. Numerous experiments are carried out to validate the effectiveness of the proposed method. Compared to traditional and state-of-the-art methods, AWFLN performs the best both subjectively and objectively, with high efficiency. The code is available at https://github.com/yotick/AWFLN.

Context Adaptive Network for Image Inpainting.

In a typical image inpainting task, the location and shape of the damaged or masked area is often random and irregular. The vanilla convolutions widely used in learning-based inpainting models treat all spatial features as valid and share parameters across regions, making it difficult for them to cope with those irregular damages, and models tend to produce inpainting results with color discrepancy and blurriness. In this paper, we propose a novel Context Adaptive Network (CANet) to address this issue. The main idea of the proposed CANet is able to generate different weights depending on the miscellaneous input, which may help to complement images with multiple broken forms in a flexible way. Specifically, the proposed CANet has two novel context adaptive modules, namely, the context adaptive block (CAB) and the cross-scale contextual attention (CSCA), which utilize attention mechanisms to cope with diverse content breakdowns. The proposed CAB, during the forward propagation, uses an adaptive term to determine the importance between adaptive term and convolution kernel, so as to dynamically balance features based on the degree of breakage (confidence level or soft mask), and the overall calculation is formulated as a classic convolution implementation with an additional attention term to describe local structure. Besides, the proposed CSCA, not only takes advantage of the contextual attention module, but also considers cross-scale information transfer to generate reasonable features for damaged areas, thus alleviating the inefficiency of the long-range modeling capability of convolutional neural networks. Qualitative and quantitative experiments show that our method performs better than state-of-the-arts, producing clearer, more coherent and visually plausible inpainting results. The code can be found at github.com/dengyecode/CANet_image_inpainting.

An adaptive block iterative process for a class of multiple sets split variational inequality problems and common fixed point problems in hilbert spaces

<p style='text-indent:20px;'>In this paper, we present extension of a class of split variational inequality problem and fixed point problem due to Lohawech et al. (J. Ineq Appl. 358, 2018) to a class of multiple sets split variational inequality problem and common fixed point problem (CMSSVICFP) in Hilbert spaces. Using the Halpern subgradient extragradient theorem of variational inequality problems, we propose a parallel Halpern subgradient extragradient CQ-method with adaptive step-size for solving the CMSSVICFP. We show that a sequence generated by the proposed algorithm converges strongly to the solution of the CMSSVICFP. We give a numerical example and perform some preliminary numerical tests to illustrate the numerical efficiency of our method.</p>

Joint grading of diabetic retinopathy and diabetic macular edema using an adaptive attention block and semisupervised learning

Joint grading of diabetic retinopathy and diabetic macular edema using an adaptive attention block and semisupervised learning

Computing Offloading and Resource Optimization of Adaptive Data Block Size in Edge Environments

Computing Offloading and Resource Optimization of Adaptive Data Block Size in Edge Environments

Pseudo-Labeling and Meta Reweighting Learning for Image Aesthetic Quality Assessment

In the tasks of image aesthetic quality assessment, it is difficult to reach both the high score area and low score area due to the normal distribution of aesthetic datasets. To reduce the error in labelling and solve the problem of normal data distribution, we propose a new aesthetic mixed dataset with classification and regression called AMD-CR, and we train a meta reweighting network to reweight the loss of training data differently. In addition, we provide a training strategy according to different stages based on pseudo labels, and then we use it for aesthetic training according to different stages in classification and regression tasks. In the construction of the network structure, we construct an aesthetic adaptive block (AAB) structure that can adapt to any size of the input images. Besides, we also use the efficient channel attention (ECA) to strengthen the feature extracting ability of each task. The experimental result shows that our method improves 0.1112 compared with the conventional method in SROCC. The method can also help to find best aesthetic path planning for unmanned aerial vehicles (UAV) and vehicles.

A fault diagnosis method of rolling bearing based on improved deep residual shrinkage networks

Aiming at the problem of signal distortion caused by deep residual shrinkage network (DRSN) in the noise reduction process, improved deep residual shrinkage network (IDRSN) are proposed and applied to rolling bearing fault diagnosis under noise backgrounds. Firstly, we design an improved pseudo-soft threshold function (IPSTF) to eliminate the signal distortion caused by the soft threshold function(STF). Then, a pseudo-soft threshold block (PSTB) and an adaptive slope block (ASB) are proposed to construct an improved residual shrinkage building unit (IRSBU) for setting the optimal threshold and slope adaptively. Finally, the method is applied to rolling bearing fault diagnosis in two different operating conditions under noise backgrounds. The results show that the proposed method has higher accuracy and robustness than the existing methods.

An efficient medical image compression technique for telemedicine systems

The medical practitioners primarily used medical images to reveal abnormalities in the internal critical organs and structures of body covered by the bones and the skin. Main application of medical imaging is to perform medical diagnosis from the image features extracted. Processing these images are very much required for assessing the patient’s condition. However, long-term monitoring of the patient using certain medical imaging technologies produces enormous volumes of data everyday. There is a need to compress the data to reduce redundancies and speed up the acquisition process, making them suitable for efficient transmission and analysis. Recently Compressed Sensing (CS) has been widely used for image compression at high speed with fewer samples. High-quality reconstruction using conventional CS and Block based CS (BCS) is a matter of utmost concern as they follow the random selection of samples. This could be overcome by adaptively selecting samples from various image regions using Adaptive Block Compressed Sensing (ABCS). This paper proposes Coefficient Mixed Thresholding based ABCS (CMT-ABCS) for compressing different medical images with a high compression ratio. The experimental outcomes exhibit a noteworthy improvement in the proposed method’s performance metrics when compared to other state-of-the-art approaches. There is a increase in PSNR of 5–10 dB, SSIM of 0.1–0.2 with NCC values closer to 1 and NAE values closer to 0. At low sampling rate, the reconstruction was greatly enhanced with only around 10% measurements/samples.

A quality enhancement network with coding priors for constant bit rate video coding

Video coding can effectively compress data, while the introduction of compression artifacts degrades the visual quality and the performance of artificial intelligence video applications. Video quality enhancement (VQE) methods can improve compressed video quality without modifying the coding standard’s main modules. The existing VQE works mainly aim to improve video quality in the constant quantization parameter (CQP) coding mode. However, constant bit rate (CBR) coding mode is widely used in some streaming playback video applications, and VQE at CBR is more challenging than that at CQP. This article presents a novel Constant bit rate Video quality enhancement Network combined with Coding priors (CVCN). The proposed CVCN can be inserted into High Efficiency Video Coding (HEVC) codec as a CNN-based in-loop filter (LF) module or a post-processing module out of the codec. Moreover, we design a two-pass training strategy for the LF module to overcome multiple filtering. To adapt the QP diversity of CBR videos, we utilize the coding unit (CU)-wise QP prior by constructing a CU-wise QP adaptive module (CQAM) and a QP adaptive multi-scale residual block (QAMSRB) based on CQAM. We construct the CU-partition prior (CPP) by exploiting the relationship between the compression noise and CU partition information. A novel CPP spatial-attention block is proposed to combine the CPP with the self-attention module. Furthermore, the proposed CVCN can effectively enhance CBR videos at different bits per pixel (BPP) via a single model. Extensive experimental results show the superiority of the CVCN over state-of-the-art VQE approaches for CBR videos.

Assessment of trunk control and core stability in spinal cord injured subjects

Objective measures for evaluating active trunk movements and postural control in sitting are important for the assessment of people with Spinal Cord Injury (SCI) and for tailoring rehabilitative interventions. However, studies addressing this issue are limited [1]. This work fills this gap by proposing a modular protocol for the kinematic and muscular quantitative assessment of trunk control, specifically designed for SCI subjects. We used hunova® (Movendo Technology), a medical robotic device for the sensory-motor assessment and training of postural control in standing and sitting [2]. consisted of two different tests: a reaching (R) and an adaptive exercise (A), both performed while sitting on hunova®. The test R aims to assess the volitional component of trunk control: six reaching exercises, where subjects need to actively tilt the seat platform to move a cursor on the screen and reach targets, are alternating with two different platform inclinations: 3.5 and 5° (1° corresponds to 1 cm of cursor motion on the screen). The test A aims to assess how subjects adapt to predictable perturbations imposed by the seat platform [3]: six repetitions of an adaptive task, where subjects need to stabilize themselves while the seat platform moves following a given infinite-shape trajectory, is repeated once with eyes open (EO) and eyes close (EC), with different inclinations of the seat platform: 3, 5 and 7°. After testing the device and protocol on 18 healthy subjects, 10 SCI subjects (with different lesion level) participated in the study. Both the reaching block and the adaptive block (Fig. 1) started with the easiest testing condition and gradually increases the difficulty of the exercises. This allowed all SCI subjects, with different lesion level, to participate in the assessment. Electromyography (EMG) was recorded bilaterally from the following muscles: obliquus externus abdominis, trapezius descendens, trapezius transversalis, trapezius ascendens, erector spinae longissimus, erector spinae iliocostalis, multifidus, latissimus dorsi, soleus, tibialis anterior, vastus medialis, semitendinosus, gluteus medium. The leg muscles were recorded only for SCI subjects with an incomplete lesion of the spinal cord. For each subject, we characterized postural responses in terms of kinematics and muscular activity, estimating the correspondent spatiotemporal organization of the motoneuronal activity in the spinal cord [4]. All subjects with an incomplete SCI could finish the whole protocol. This was not the case for subjects with a complete lesion. Only one subject (lesion level D12) completed both blocks and had movement performance like SCI subjects with an incomplete lesion. The other subjects (two with lesion level D6 and one D3) performed few repetitions of R as they found difficult to control the seat platform, and the tasks with 3 and 5° of inclination for A. Concerning the performance, for R all subjects, independently on their type of lesion, reached a greater number of targets across repetitions. During A, all subjects had a bigger trunk sway for higher inclinations of the seat platform, with higher values in the repetitions with EC. Each SCI subject used his/her own muscular strategy for reaching the different targets and also for adapting to the seat motion, despite it was continuous and predictable. This assessment highlighted the differences in the muscular control among subjects, also due to the different lesions. This modular protocol evaluated in a quantitative, objective, repeatable, and comprehensive manner the kinematic and muscular strategies adopted by SCI subjects when performing dynamic tasks while sitting. The device hunova® allowed us to evaluate different aspects related to trunk control, the volitional and the feedforward control of balance while adapting to predictable and continuous perturbations. The results of this assessment will be used to tailor specific rehabilitative interventions.

Blind-noise image denoising with block-matching domain transformation filtering and improved guided filtering

The adaptive block size processing method in different image areas makes block-matching and 3D-filtering (BM3D) have a very good image denoising effect. Based on these observation, in this paper, we improve BM3D in three aspects: adaptive noise variance estimation, domain transformation filtering and nonlinear filtering. First, we improve the noise-variance estimation method of principle component analysis using multilayer wavelet decomposition. Second, we propose compressive sensing based Gaussian sequence Hartley domain transform filtering to reduce noise. Finally, we perform edge-preserving smoothing on the preprocessed image using the guided filtering based on total variation. Experimental results show that the proposed denoising method can be competitive with many representative denoising methods on the evaluation criteria of PSNR. However, it is worth further research on the visual quality of denoised images.

Lightweight spatial-channel adaptive coordination of multilevel refinement enhancement network for image reconstruction

Benefiting from vigorous development of the deep learning, many CNN-based image super-resolution methods have been appearing and achieve better results than traditional algorithms. However, it is difficult for most algorithms to adaptively adjust spatial region and channel features at the same time, let alone the information exchange between them. In addition, the exchange of information between attention modules is even less visible to researchers. To solve these problems, we put forward a lightweight spatial-channel adaptive coordination of multilevel refinement enhancement network(MREN). Specifically, we construct a space-channel adaptive coordination block, which enables the network to learn the spatial region and channel feature information of interest under different receptive fields. Furthermore, the information of corresponding feature processing level in the spatial part and the channel part is exchanged by jumping connection to achieve the coordination between the two. We built a bridge of communication between attention modules through a simple linear combination operation, so as to more accurately and continuously guide the network to pay attention to the information of interest. Extensive experiments on several standard test sets have shown that our MREN achieves superior performance over other advanced algorithms with a very small number of parameters and very low computational complexity.

UStark: underwater image domain-adaptive tracker based on Stark

Though the open-air tracker has achieved an advanced level, its design is still a challenging task for degraded underwater images. Using underwater enhancement technology can improve the performance of underwater trackers. However, most underwater image enhancement methods focus on improving the visual effect rather than serving the tracker better. Therefore, we intended to explore a simple but powerful image domain-adaptive method to improve Stark’s performance by enhancing Stark’s input images. Specifically, it consists of underwater image adaptation network (UIAN) with double heads and adaptation block based on scene estimation (ABSE) that consists of three independent image processing modules without deep learning. UIAN is used to predict the category of image domain and parameters of ABSE. ABSE decodes the parameters and sequentially process underwater images in each module. The training of UIAN is independent of the training of the tracker. After training the class prediction head of UIAN first, freezing its weights, by initializing and tracking in one enhanced image and computing tracker’s loss, the parameter head can be trained to make sure UIANs hyperparameters can match the Stark tracker. The UStark proposed can adaptively process clear and degraded underwater images. Compared with Stark, UStark has improved the accuracy and success rate in typical underwater environments by 3.7% and 1.5% (blue), 5% and 3.4% (yellow), and 5.4% and 3.3% (dark), respectively. In addition, compared with other underwater image enhancement methods, our method can enhance the performance of the tracker in more categories of underwater images.

Adaptive resize-residual deep neural network for fault diagnosis of rotating machinery

Accurate fault diagnosis technology is essential for ensuring reliable operation of rotating machinery. However, complex conditions and various damage forms bring challenges to present diagnosis technology. In this study, a novel fault diagnosis method is proposed utilizing the newly developed adaptive resize-residual deep neural networks. The usage of the proposed method consists of three steps. First, the continuous wavelet transform is used to transfer the acquired vibration signals into time–frequency images. Second, the histogram equalization algorithm is applied to enhance the contrast of these images. Finally, the enhanced images are used as the input of newly proposed adaptive resize-residual networks, in which the adaptive resize block can deduce the dimensions of input data by self-learning and feed them into the residual block for pattern recognition. Two experimental cases are designed to evaluate the performance of proposed method. The experimental results indicate that the proposed adaptive resize-residual network obtains superior recognition accuracy and outperforms many state-of-the-art methods.

MEA-Net: A Lightweight SAR Ship Detection Model for Imbalanced Datasets

The existing synthetic aperture radar (SAR) ship datasets have an imbalanced number of inshore and offshore ship targets, and the number of small, medium and large ship targets differs greatly. At the same time, the existing SAR ship detection models in the application have a huge structure and require high computing resources. To solve these problems, we propose a SAR ship detection model named mask efficient adaptive network (MEA-Net), which is lightweight and high-accuracy for imbalanced datasets. Specifically, we propose the following three innovative modules. Firstly, we propose a mask data balance augmentation (MDBA) method, which solves the imbalance of sample data between inshore and offshore ship targets by combining mathematical morphological processing and ship label data to greatly improve the ability of the model to detect inshore ship targets. Secondly, we propose an efficient attention mechanism (EAM), which effectively integrates channel features and spatial features through one-dimensional convolution and two-dimensional convolution, to improve the feature extraction ability of the model for SAR ship targets. Thirdly, we propose an adaptive receptive field block (ARFB), which can achieve more effective multi-scale detection by establishing the mapping relationship between the size of the convolution kernel and the channel of feature map, to improve the detection ability of the model for ship targets of different sizes. Finally, MEA-Net is deployed on the Jeston Nano edge computing device of the 2 GB version. We conducted experimental validation on the SSDD and HRSID datasets. Compared with the baseline, the AP of MEA-Net increased by 2.18% on the SSDD dataset and 3.64% on the HRSID dataset. The FLOPs and model parameters of MEA-Net were only 2.80 G and 0.96 M, respectively. In addition, the FPS reached 6.31 on the Jeston Nano, which has broad application prospects.

Theoretical basis of the formation of adaptive management of enterprise financial security

Methods. The research is based on the use of such methods as abstraction – when defining the essence of the concept of «financial security of the enterprise», analysis and synthesis – when forming the principles of the process of ensuring the financial security of the enterprise, a systems approach – when identifying the elements of adaptive management of the financial security of the enterprise. Results. The article substantiates the theoretical foundations of the formation of adaptive management of the financial security of the enterprise. It has been demonstrated that the financial security of the enterprise must be determined from the point of view of an integrated approach, taking into account the financial condition characterized by the appropriate level of balance of financial resources, tools and technologies; resistance to the influence of internal and external environmental factors; the appropriate level of efficiency and development of the financial system of the enterprise with sufficient amounts of financial resources; a state that provides for the security of the enterprise’s financial resources in the context of its economic relations in current and strategic perspectives. Novelty. The basic principles on which the financial security of the enterprise is based are defined, with the internal and external threats to the financial security of the enterprise being singled out. The main elements of adaptive management of the enterprise's financial security are grouped into preparatory and informational, functional and monitoring, and control and adaptation blocks. Practical value. On the basis of the conducted research, it was determined that proper and timely monitoring of the enterprise operating environment in order to qualitatively identify and determine the impact of threats on financial security is an important stage of the adaptive management of the financial security of the enterprise, which involves the development and selection of adaptive protective actions aimed at neutralizing the negative impact on the financial the state of the enterprise and the sustainability of its economic development.