Neural Network For Image Restoration Research Articles

A Global Self-Attention Memristive Neural Network for Image Restoration

A Global Self-Attention Memristive Neural Network for Image Restoration

Mixup-Based Neural Network for Image Restoration and Structure Prediction From SEM Images

Mixup-Based Neural Network for Image Restoration and Structure Prediction From SEM Images

MWA-MNN: Multi-patch Wavelet Attention Memristive Neural Network for image restoration

Adverse weather conditions can severely reduce the image quality captured by outdoor imaging devices. However, most existing image restoration algorithms are designed for specific tasks, lack generalization capabilities, and are unable to balance spatial details with contextual feature. Moreover, outdoor images are typically captured using edge devices with limited computing resources, rendering it challenging to deploy image restoration algorithms on edge devices. To tackle these problems, we propose a memristive-based image restoration algorithm, named Multi-patch Wavelet Attention Memristive Neural Network (MWA-MNN). Multiple image restoration tasks can be achieved by implementing Convolutional Neural Network (CNN) based on memristor crossbar arrays, as well as specified memristive circuits. Our proposed network comprises three sub-networks with multi-patch design enabling each sub-network to focus on different scales, thereby preserving more spatial details while fusing rich contextual feature. We propose the Discrete Wavelet Attention Block (DWAB) and the Coordinate Channel Attention Block (CCAB) to extract rich contextual information and the Supervised Selective Kernel Block (SSKB) to effectively filter the extracted features. Our proposed memristor-based circuit implementation provides an effective solution for deploying image restoration algorithms on edge devices. Experiment results on multiple benchmark datasets for image deraining, dehazing, and low-light image enhancement show that our proposed method outperforms over 20 state-of-the-art methods.

A Variational neural network for image restoration based on coupled regularizers

A Variational neural network for image restoration based on coupled regularizers

On quantization of convolutional neural networks for image restoration

Recently, commercial vision sensors hit the mobile market. To achieve that, computer vision networks had to be quantized. However, this topic was not studied well for Image Signal processor (ISP) challenging image restoration tasks, being crucially important for hardware implementation, as well as for deployment on hardware accelerators, e.g. Neural Processors Units (NPU). In this paper, we studied the effect of the quantization of deep learning network on image quality. We tried various quantization on raw RGBW image demosaicing. Experimental results show that 12 bit weight quantization can sustain image quality at the similar level with floating-point network. 10 bit quantized network shows slight degradation in objective image quality and mild visual artifacts. If network weight’s bit-depth can be significantly reduced for computer vision tasks, our finding shows that it is not true for raw image restoration tasks: at least 10 bit weights are required to provide sufficient image quality. However, we can save some memory on feature maps bit-depth. We can conclude that network bit depth is critical for raw image restoration.

Removal of speckle noises from ultrasound images using five different deep learning networks

Image enhancement methods are applied to medical images to reduce the noise that they contain. There are many academic studies in the literature using classical image enhancement methods. Ultrasound imaging is a medical imaging method that is used for the diagnosis of diseases. In this study, speckle noises with Rayleigh distribution at four different noise levels (σ = 0.10, 0.25, 0.50, 0.75) are added to ultrasound images of the brachial plexus nerve region. Five different deep learning networks (Dilated Convolution Autoencoder Denoising Network/Di-Conv-AE-Net, Denoising U-Shaped Net/D-U-Net, BatchRenormalization U-Net/Br-U-Net, Generative Adversarial Denoising Network/DGan-Net, and CNN Residual Network/DeRNet) are used for reducing the speckle noises of the ultrasound images. The performances of the deep networks are compared with block-matching and 3D filtering (BM3D), which is one of the most preferred classical image enhancement algorithms; with classical filters including Bilateral, Frost, Kuan, Lee, Mean, and Median Filters; and with deep learning networks including Learning Pixel-Distribution Prior with Wider Convolution for Image Denoising (WIN5-RB), Denoising Prior Driven Deep Neural Network for Image Restoration (DPDNN), and Fingerprint Image Denoising and Inpainting Using M-Net Based Convolutional Neural Networks (FPD-M-Net). Network performance is evaluated according to peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), and runtime criteria and the proposed deep learning networks are shown to outperform the other networks.

CANet: Concatenated Attention Neural Network for Image Restoration

In this article, we present a general framework for low-level vision tasks including image compression artifacts reduction and image denoising. Under this framework, a novel concatenated attention neural network (CANet) is specifically designed for image restoration. The main contributions of this article are as follows: First, we establish a concise network with a recursive topology by abandoning most complex artificially designed topology of the network. Second, we did not use the down-sampling mechanism in our network and a lightweight attention mechanism is adopted for this. Third, we rethink the feature fusion mechanism in the field of image restoration and improve this by applying concise but effective concatenation and feature selection mechanism which promotes further extraction of more cleaner features in images. Lastly, we demonstrate that CANet achieves better results than previous state-of-the-art approaches with sufficient experiments in compression artifacts removing and image denoising.

Alternative continuous- and discrete-time neural networks for image restoration

ABSTRACTThis paper presents alternative continuous- and discrete-time neural networks for image restoration in real time by introducing new vectors and transforming its optimization conditions into a system of double projection equations. The proposed neural networks are shown to be stable in the sense of Lyapunov and convergent for any starting point. Compared with the existing neural networks for image restoration, the proposed models have the least neurons, a one-layer structure and the faster convergence, and is suitable to parallel implementation. The validity and transient behaviour of the proposed neural network is demonstrated by numerical examples.

Content-aware image restoration for electron microscopy.

Multiple approaches to use deep neural networks for image restoration have recently been proposed. Training such networks requires well registered pairs of high and low-quality images. While this is easily achievable for many imaging modalities, e.g., fluorescence light microscopy, for others it is not. Here we summarize on a number of recent developments in the fast-paced field of Content-Aware Image Restoration (CARE), in particular, and the associated area of neural network training, more in general. We then give specific examples how electron microscopy data can benefit from these new technologies.

Denoising Prior Driven Deep Neural Network for Image Restoration.

Deep neural networks (DNNs) have shown very promising results for various image restoration (IR) tasks. However, the design of network architectures remains a major challenging for achieving further improvements. While most existing DNN-based methods solve the IR problems by directly mapping low quality images to desirable high-quality images, the observation models characterizing the image degradation processes have been largely ignored. In this paper, we first propose a denoising-based IR algorithm, whose iterative steps can be computed efficiently. Then, the iterative process is unfolded into a deep neural network, which is composed of multiple denoisers modules interleaved with back-projection (BP) modules that ensure the observation consistencies. A convolutional neural network (CNN) based denoiser that can exploit the multi-scale redundancies of natural images is proposed. As such, the proposed network not only exploits the powerful denoising ability of DNNs, but also leverages the prior of the observation model. Through end-to-end training, both the denoisers and the BP modules can be jointly optimized. Experimental results on several IR tasks, e.g., image denoisig, super-resolution and deblurring show that the proposed method can lead to very competitive and often state-of-the-art results on several IR tasks, including image denoising, deblurring, and super-resolution.

Image Restoration by Using New AGA Optimized BPNN

In this paper a new method of Adaptive Genetic Algorithm (AGA) is introduced to optimize the back propagation neural network for Image restoration. In this new AGA, we added permutation operator in addition to traditional mutation and Pooling Operator is introduced. To increase the convergence rate, we used adaptive crossover rate and mutation rate. It has been observed that with the addition of these two operators, the use of Genetic Algorithms (GA) for navigating the optimal combination of solution is more effective and the convergence can be achieved with more accuracy. In addition, it also decreases the value of Mean Square Error (MSE) significantly.

Estimate of Convergence Rate of Neural Networks for Image Restoration

This paper deals with neural networks of 8-neighbor for image restoration which are used for gray images. Generally, since comparatively long operation time is needed when a neural network is applied to a certain problem, it is important to estimate the convergence rate of the neural network in order to evaluate the operation time. In this paper, the theoretical convergence rate of the neural network for image restoration is derived both for the case where an image consists of not many pixels and for the case where an image consists of many pixels. Furthermore, it is shown that the two convergence rates are appropriate through a numerical experiment.

画像復元用ニューラルネットワークの安定性について

画像復元用ニューラルネットワークの安定性について

Improvement on performance of modified Hopfield neural network for image restoration

By analyzing the same inequality ||u*||(1)=/<(1/2)trace(T), the authors conclude that a severely blurred image is generally restored less accurately than a mildly blurred one by the modified Hopfield neural network. This conclusion is the opposite of the statement made in Paik and Katsaggelos (1992). The authors also propose an improved new algorithm. Simulation results show that the SNRs of the images restored by the algorithm are higher by 3 to 8 db than those restored by the algorithm in Paik and Katsaggelos and the streaks in the restored images are obviously suppressed by the algorithm.

Sequential and parallel image restoration: neural network implementations

Sequential and parallel image restoration algorithms and their implementations on neural networks are proposed. For images degraded by linear blur and contaminated by additive white Gaussian noise, maximum a posteriori (MAP) estimation and regularization theory lead to the same high dimension convex optimization problem. The commonly adopted strategy (in using neural networks for image restoration) is to map the objective function of the optimization problem into the energy of a predefined network, taking advantage of its energy minimization properties. Departing from this approach, we propose neural implementations of iterative minimization algorithms which are first proved to converge. The developed schemes are based on modified Hopfield (1985) networks of graded elements, with both sequential and parallel updating schedules. An algorithm supported on a fully standard Hopfield network (binary elements and zero autoconnections) is also considered. Robustness with respect to finite numerical precision is studied, and examples with real images are presented.