In the registration task, the transformation should ideally output a deformation field with diffeomorphic properties, that is, the deformation field is reversible and smooth so that the topology of the image will not ...
详细信息
In the registration task, the transformation should ideally output a deformation field with diffeomorphic properties, that is, the deformation field is reversible and smooth so that the topology of the image will not change during the transformation. However, the performance of existing registration methods in ensuring the smoothness and reversibility of the deformation field is still unsatisfactory. In this paper, a novel inverse consistency neural field (ICNF) method was proposed, which can guarantee the reversibility of the registration transformation and significantly improve the regularity of the registration deformation field. The proposed method is based on pairwise optimization and takes advantage of the powerful representational capabilities of deep neural networks to model the transformations between the joint symmetric estimation image pairs. Based on the different methods of generating deformation fields, the proposed method can either directly output forward and backward displacement fields, realized as displacement field deformable image registration (ICNF- disp), or generate forward and backward time-dependent velocity fields, and integrate these velocity fields to derive the deformation fields, known as diffeomorphic deformable image registration (ICNF-diff ). To reduce the impact of interpolation errors in the generated velocity field on ICNF-diff registration performance, we propose a novel inverse consistent loss function for the velocity field. By imposing an inverse consistency constraint on the time-dependent velocity vector, the invertibility and topological preservation of the transformation are further ensured. Extensive experiments on a public Magnetic Resonance 3D brain scan dataset show that the proposed method guarantees invertibility of the transformation between image pairs while outperforming the state-of-the-art registration methods on registration regularity (ICNF-disp improved 86.03% and ICNF-diff improved 97.12%).
Computed tomography (CT) is one of the most important medical imaging technologies in use today. Most commercial CT products use a technique known as the filtered backprojection (FBP) that is fast and can produce dece...
详细信息
Computed tomography (CT) is one of the most important medical imaging technologies in use today. Most commercial CT products use a technique known as the filtered backprojection (FBP) that is fast and can produce decent image quality when an X-ray dose is high. However, the FBP is not good enough on low-dose X-ray CT imaging because the CT image reconstruction problem becomes more stochastic. A more effective reconstruction technique proposed recently and implemented in a limited number of CT commercial products is an iterative reconstruction (IR). The IR technique is based on a Bayesian formulation of the CT image reconstruction problem with an explicit model of the CT scanning, including its stochastic nature, and a prior model that incorporates our knowledge about what a good CT image should look like. However, constructing such prior knowledge is more complicated than it seems. In this article, we propose a novel neural network for CT image reconstruction. The network is based on the IR formulation and constructed with a recurrent neural network (RNN). Specifically, we transform the gated recurrent unit (GRU) into a neural network performing CT image reconstruction. We call it ``GRU reconstruction.'' This neural network conducts concurrent dual-domain learning. Many deep learning (DL)-based methods in medical imaging are single-domain learning, but dual-domain learning performs better because it learns from both the sinogram and the image domain. In addition, we propose backpropagation through stage (BPTS) as a new RNN backpropagation algorithm. It is similar to the backpropagation through time (BPTT) of an RNN;however, it is tailored for iterative optimization. Results from extensive experiments indicate that our proposed method outperforms conventional model-based methods, single-domain DL methods, and state-of-the-art DL techniques in terms of the root mean squared error (RMSE), the peak signal-to-noise ratio (PSNR), and the structure similarity (SSIM) and in
We propose a novel approach to enhance image demosaicking algorithms using implicit neural representations (INR). Our method employs a multi-layer perceptron to encode RGB images, combining original Bayer measurements...
详细信息
We propose a novel approach to enhance image demosaicking algorithms using implicit neural representations (INR). Our method employs a multi-layer perceptron to encode RGB images, combining original Bayer measurements with an initial estimate from existing demosaicking methods to achieve superior reconstructions. A key innovation is the integration of two loss functions: a Bayer loss for fidelity to sensor data and a complementary loss that regularizes reconstruction using interpolated data from the initial estimate. This combination, along with INR's inherent ability to capture fine details, enables high-fidelity reconstructions that incorporate information from both sources. Furthermore, we demonstrate that INR can effectively correct artifacts in state-of-the-art demosaicking methods when input data diverge from the training distribution, such as in cases of noise or blur. This adaptability highlights the transformative potential of INR-based demosaicking, offering a robust solution to this challenging problem.
Rain streaks typically cause significant visual degradation and foreground occlusions, hindering the progress of visual tasks in outdoor scenarios. Existing image deraining methods, predominantly based on Convolutiona...
详细信息
Rain streaks typically cause significant visual degradation and foreground occlusions, hindering the progress of visual tasks in outdoor scenarios. Existing image deraining methods, predominantly based on Convolutional neural Networks (CNNs), exhibit certain limitations. These methods tend to overly focus on low-level visual features, demonstrating insufficient ability to capture high-dimensional global features. Furthermore, they often lack targeted attention to channel information and spatial details, which restricts their effectiveness. To address these shortcomings, this paper proposes the Delta-Calibration Derain Network (DCD-Net). The DCD-Net introduces a sequential Delta Convolutional Layer structure to significantly expand the feature acquisition range. Additionally, this study pioneers the Joint Calibration Attention module, which precisely captures both channel and spatial feature information, leading to enhanced network performance. Experimental results across multiple synthetic datasets show that the proposed method achieves superior performance in terms of Peak signal-to-Noise Ratio and Structural Similarity Index, validating the advantages of DCD-Net over traditional CNN-based models.
The Hurst exponent is used to identify the autocorrelation structure of a stochastic time series, which allows for detecting persistence in time series data. Traditional signalprocessing techniques work reasonably we...
详细信息
The Hurst exponent is used to identify the autocorrelation structure of a stochastic time series, which allows for detecting persistence in time series data. Traditional signalprocessing techniques work reasonably well in determining the Hurst exponent of a stochastic time series. However, a notable drawback of these methods is their speed of computation. neural networks have repeatedly proven their ability to learn very complex input-output mappings, even in high dimensional vector spaces. Therefore, an endeavour has been undertaken to employ neural networks to determine the Hurst exponent of a stochastic time series. Unlike previous attempts to solve such problems using neural networks, the proposed architecture can be recognised as the universal estimator of Hurst exponent for short-range and long-range dependent stochastic time series. Experiments demonstrate that if sufficiently trained, neural network can predict the Hurst exponent of any stochastic data at least fifteen times faster than standard signalprocessing approaches.
To address the problem that traditional convolutional neural networks cannot classify facial expression image features precisely, an interpretable face expression recognition method combining ResNet18 residual network...
详细信息
To address the problem that traditional convolutional neural networks cannot classify facial expression image features precisely, an interpretable face expression recognition method combining ResNet18 residual network and support vector machines (SVM) is proposed in the paper. The SVM classifier is used to enhance the matching ability of feature vectors and labels under the expression image feature space, to improve the expression recognition effect of the whole model. The class activation mapping and t-distributed stochastic neighbor embedding methods are used to visualize and interpret facial expression recognition's feature analysis and decision making under the residual neural network. The experimental results and the interpretable visualization analysis show that the algorithm structure can effectively improve the recognition ability of the network. Under the FER2013, JAFFE, and CK+ datasets, it achieved 67.65%, 84.44%, and 96.94% emotional recognition accuracy, respectively, showing a certain generalization ability and superior performance.
stochastic gradient descent with momentum (SGDM) is a classic optimization method that determines the update direction using a moving average of the gradient over historical steps. However, SGDM suffers from slow conv...
详细信息
stochastic gradient descent with momentum (SGDM) is a classic optimization method that determines the update direction using a moving average of the gradient over historical steps. However, SGDM suffers from slow convergence. In 2022, Yuan et al. [6] proposed stochastic gradient descent with momentum and difference (SGDMD), which incorporates the concept of differences to adjust the convergence direction and accelerate the optimization process. Despite its improvements, SGDMD requires careful parameter tuning and is prone to oscillations due to the difference mechanism. In this work, we introduce a new momentum method: stochastic gradient descent with adaptive momentum and difference (SGDAMD). Compared to SGDMD, SGDAMD demonstrates superior performance in experiments, achieving greater stability in terms of both loss values and accuracy in deep learning image classification tasks. Additionally, SGDAMD attains a sublinear convergence rate in non-convex settings while requiring less restrictive assumptions than standard smoothness conditions. These features underscore the algorithm's efficiency and effectiveness in addressing complex optimization challenges.
image denoising is a fundamental task in imageprocessing and low-level computer vision, often necessitating a delicate balance between noise removal and the preservation of fine details. In recent years, deep learnin...
详细信息
image denoising is a fundamental task in imageprocessing and low-level computer vision, often necessitating a delicate balance between noise removal and the preservation of fine details. In recent years, deep learning approaches, particularly those utilizing various neural network architectures, have shown significant promise in addressing this challenge. In this study, we propose DuINet, a novel dual-branch network specifically designed to capture complementary aspects of image information. DuINet integrates an information exchange module that facilitates effective feature sharing between the branches, and it incorporates a perceptual loss function aimed at enhancing the visual quality of the denoised images. Extensive experimental results demonstrate that DuINet surpasses existing dual-branch models and several state-of-the-art convolutional neural network (CNN)- based methods, particularly under conditions of severe noise where preserving fine details and textures is critical. Moreover, DuINet maintains competitive performance in terms of the LPIPS index when compared to deeper or larger networks such as Restormer and MIRNet, underscoring its ability to deliver high visual quality in denoised images.
S-boxes are essential in image encryption, particularly in block ciphers, but selecting the best one is challenging due to the complexity and uncertainty involved in evaluating multiple criteria. Traditional methods o...
详细信息
S-boxes are essential in image encryption, particularly in block ciphers, but selecting the best one is challenging due to the complexity and uncertainty involved in evaluating multiple criteria. Traditional methods often struggle to handle this complexity, making the decision process unclear and unreliable. We evaluate 8 x 8 S-boxes by applying them to two images and analyzing their performance across multiple criteria. Since each criterion suggests a different S-box, the selection process becomes confusing, reflecting a real-world challenge in encryption security. To address this challenge, we propose an intelligent decision-making model based on linguistic neural networks to select the most suitable S-box for image encryption. The proposed model consistently identifies Skipjack as the most effective S-box for image encryption. To ensure its accuracy and reliability, we compare it with existing MCDM models using statistical analysis and performance ranking. The results consistently confirm Skipjack as the best choice, demonstrating that the proposed model enhances security, improves decision-making, and strengthens encryption efficiency.
In conditions of multi-fault coupling, varying loads and speeds, as well as noise interference, bearing vibration signals present various complex issues, leading to difficulties in feature extraction and the need for ...
详细信息
In conditions of multi-fault coupling, varying loads and speeds, as well as noise interference, bearing vibration signals present various complex issues, leading to difficulties in feature extraction and the need for a large number of training samples for diagnostic methods. This paper designs a multi-fault coupling experiment for rolling bearings under varying load and speed conditions and proposes a new fault diagnosis method that uses the power spectrum of the AR model and a convolutional neural network to diagnose complex multi-faults in rolling bearings. It takes the original vibration signal as input, uses the AR model to convert the time-domain signal into a power spectrum, and then classifies it using a convolutional neural network. To test the performance of the AR model power spectrum convolutional neural network, this method was compared with some fault diagnosis methods. The results show that this method can achieve higher diagnostic accuracy under varying loads and speeds, and requires fewer training samples. In addition, the noise resistance of this method is also superior to other fault diagnosis methods.
暂无评论