基于PCAU-Net的快速多通道磁共振成像方法

多通道磁共振成像方法采用多个接收线圈同时欠采样k空间以加快成像速度,并基于后处理算法重建图像,但在较高加速因子时,其图像重建质量仍然较差．本文提出了一种基于PCAU-Net的快速多通道磁共振成像方法,将单通道实数U型卷积神经网络拓展到多通道复数卷积神经网络,设计了一种结构不对称的U型网络结构,通过在解码部分减小网络规模以降低模型的复杂度．PCAU-Net网络在跳跃连接前增加了1×1卷积,以实现跨通道信息交互．输入和输出之间利用残差连接为误差的反向传播提供捷径．实验结果表明,使用规则和随机采样模板,在不同加速因子时,相比常规的GRAPPA重建算法和SPIRiT重建方法,本文提出的PCAU-Net方法可高质量重建出磁共振复数图像,并且相比于PCU-Net方法,PCAU-Net减少了模型参数、缩短了训练时间．     

多局部残差连接注意网络的图像去模糊

针对现有的基于卷积神经网络的图像去模糊算法存在图像纹理细节恢复不清晰的问题,提出了一种基于多局部残差连接注意网络的图像去模糊算法。首先,采用一个卷积层进行浅层特征提取;其次,设计了一种新的基于残差连接和并行注意机制的多局部残差连接注意模块,用于消除图像模糊并提取上下文信息;再次,采用一个基于扩张卷积的成对连接模块进行细节恢复;最后,利用一个卷积层重建清晰图像。实验结果表明：在GoPro数据集上的PSNR (peak signal to noise ratio)和SSIM (structure similarity)分别为31.83 dB、0.927 5,在定性和定量两方面都表明所提方法能够有效地恢复模糊图像的纹理细节,网络性能优于对比方法。     

MRI denoising using progressively distribution-based neural network

Magnetic Resonance (MR) images often suffer from noise pollution during image acquisition and transmission, which limits the accuracy of quantitative measurements from the data. Noise in magnitude MR images is usually governed by Rician distribution, due to the existence of uncorrelated Gaussian noise with zero-mean and equal variance in both the real and imaginary parts of the complex K-space data. Different from the existing MRI denoising methods that utilizing the spatial neighbor information around the pixels or patches, this work turns to capture the pixel-level distribution information by means of supervised network learning. A progressive network learning strategy is proposed via fitting the distribution of pixel-level and feature-level intensities. The proposed network consists of two residual blocks, one is used for fitting pixel domain without batch normalization layer and another one is applied for matching feature domain with batch normalization layer. Experimental results under synthetic, complex-valued and clinical MR brain images demonstrate great potential of the proposed network with substantially improved quantitative measures and visual inspections.     

门控卷积神经网络多通路声学回声消除算法*

多通路声重放系统能够增强听者的现实感与空间感,但在免提通信条件下,其不可避免会受到噪声和回声干扰,严重影响通信质量。针对上述问题,本文提出了一种基于门控卷积循环神经网络的多通路声学回声消除和噪声抑制方法。该方法以传声器接收信号和重放声道的压缩复数谱为网络输入,以近端语音的压缩复数谱为网络的输出目标,直接从传声器拾取信号中恢复近端纯净语音,无需对声重放信号进行去相关处理,解决了传统自适应滤波方法中存在的非唯一解问题,同时保证了多通路声重放质量。仿真和真实声学环境实验均表明本文所提出的方法可显著消除多通路声重放系统的噪声和回声,在语音质量和回声返回衰减增益方面均优于传统算法。     

不均匀光照和重影的仪表图像二值化方法

高压计量仪表识别过程中需要对图像进行二值化操作,然而拍摄的仪表图像多出现光照不均和数字重影现象,导致传统方法对仪表图像的二值化困难。为此,提出一种基于卷积神经网络的二值化方法,用于对复杂光照下含数字重影的仪表图像二值化。该网络使用的数据集为真实环境下的仪表图像,首先对输入的图像进行降维提取特征,然后反卷积重建图像前景,最后输出二值图。将设计的网络与传统的二值化方法进行对比,实验结果表明,经该网络训练得到的二值图数字清晰且无重影,且测得的交并比（IoU）平均值为95.12,与样本标签图像的相似度最高,能够有效解决复杂环境下有重影的仪表图像二值化问题。     

Incorporating prior knowledge via volumetric deep residual network to optimize the reconstruction of sparsely sampled MRI

For sparse sampling that accelerates magnetic resonance (MR) image acquisition, non-linear reconstruction algorithms have been developed, which incorporated patient specific a prior information. More generic a prior information could be acquired via deep learning and utilized for image reconstruction. In this study, we developed a volumetric hierarchical deep residual convolutional neural network, referred to as T-Net, to provide a data-driven end-to-end mapping from sparsely sampled MR images to fully sampled MR images, where cartilage MR images were acquired using an Ultra-short TE sequence and retrospectively undersampled using pseudo-random Cartesian and radial acquisition schemes. The network had a hierarchical architecture that promoted the sparsity of feature maps and increased the receptive field, which were valuable for signal synthesis and artifact suppression. Relatively dense local connections and global shortcuts were established to facilitate residual learning and compensate for details lost in hierarchical processing. Additionally, volumetric processing was adopted to fully exploit spatial continuity in three-dimensional space. Data consistency was further enforced. The network was trained with 336 three-dimensional images (each consisting of 32 slices) and tested by 24 images. The incorporation of a priori information acquired via deep learning facilitated high acceleration factors (as high as 8) while maintaining high image fidelity (quantitatively evaluated using the structural similarity index measurement). The proposed T-Net had an improved performance as compared to several state-of-the-art networks.     

Transfer learning in deep neural network based under-sampled MR image reconstruction

In Magnetic Resonance Imaging (MRI), the success of deep learning-based under-sampled MR image reconstruction depends on: (i) size of the training dataset, (ii) generalization capabilities of the trained neural network. Whenever there is a mismatch between the training and testing data, there is a need to retrain the neural network from scratch with thousands of MR images obtained using the same protocol. This may not be possible in MRI as it is costly and time consuming to acquire data. In this research, a transfer learning approach i.e. end-to-end fine tuning is proposed for U-Net to address the data scarcity and generalization problems of deep learning-based MR image reconstruction. First the generalization capabilities of a pre-trained U-Net (initially trained on the human brain images of 1.5 T scanner) are assessed for: (a) MR images acquired from MRI scanners of different magnetic field strengths, (b) MR images of different anatomies and (c) MR images under-sampled by different acceleration factors. Later, end-to-end fine tuning of the pre-trained U-Net is proposed for the reconstruction of the above-mentioned MR images (i.e. (a), (b) and (c)). The results show successful reconstructions obtained from the proposed method as reflected by the Structural SIMilarity index, Root Mean Square Error, Peak Signal-to-Noise Ratio and central line profile of the reconstructed images.     

Multi-Focus Image Fusion Based on Multi-Scale Generative Adversarial Network

The methods based on the convolutional neural network have demonstrated its powerful information integration ability in image fusion. However, most of the existing methods based on neural networks are only applied to a part of the fusion process. In this paper, an end-to-end multi-focus image fusion method based on a multi-scale generative adversarial network (MsGAN) is proposed that makes full use of image features by a combination of multi-scale decomposition with a convolutional neural network. Extensive qualitative and quantitative experiments on the synthetic and Lytro datasets demonstrated the effectiveness and superiority of the proposed MsGAN compared to the state-of-the-art multi-focus image fusion methods.     

基于红外和可见光图像逐级自适应融合的场景深度估计

从图像中恢复场景的深度是计算机视觉领域中的一个关键问题。考虑到单一类型图像在深度估计中受场景不同光照的限制,提出了基于红外和可见光图像逐级自适应融合的场景深度估计方法（PF-CNN）。该方法包括双流滤波器部分耦合网络、自适应多模态特征融合网络以及自适应逐级特征融合网络。在双流卷积中红外和可见光图像的滤波器部分耦合使两者特征得到增强;自适应多模态特征融合网络学习红外和可见光图像的残差特征并将两者自适应加权融合,充分利用两者的互补信息;逐级特征融合网络学习多层融合特征的结合,充分利用不同卷积层的不同特征。实验结果表明:PF-CNN在测试集上获得了较好的效果,将阈值指标提高了5%,明显优于其他方法。     

基于卷积神经网络的焊缝缺陷图像分类研究

为有效地对焊缝缺陷进行分类,从而判断焊接质量的等级,对传统卷积神经网络进行改进,提出一种多尺度压缩激励网络模型（SINet）。将4组两两串联的3×3卷积模块与Inception模块、压缩激励模块（SE block）相结合。通过多尺度压缩激励模块（SI module）将卷积层中的特征进行多尺度融合和特征重标定以提高分类准确率,并用全局平均池化层代替全连接层减少模型参数。此外考虑到焊接缺陷数量不平衡对准确率的影响,采用深度卷积对抗生成网络（DCGAN）进行数据集的平衡处理,并在该数据集上验证模型的有效性。与传统卷积神经网络相比,该模型具有良好的性能,在测试集上准确率达到96.77%,同时模型的参数个数也明显减少。结果表明该方法对焊缝缺陷图像能进行有效地分类。     

Dynamic pixel-wise weighting-based fully convolutional neural networks for left ventricle segmentation in short-axis MRI

Left ventricle (LV) segmentation in cardiac MRI is an essential procedure for quantitative diagnosis of various cardiovascular diseases. In this paper, we present a novel fully automatic left ventricle segmentation approach based on convolutional neural networks. The proposed network fully takes advantages of the hierarchical architecture and integrate the multi-scale feature together for segmenting the myocardial region of LV. Moreover, we put forward a dynamic pixel-wise weighting strategy, which can dynamically adjust the weight of each pixel according to the segmentation accuracy of upper layer and force the pixel classifier to take more attention on the misclassified ones. By this way, the LV segmentation performance of our method can be improved a lot especially for the apical and basal slices in cine MR images. The experiments on the CAP database demonstrate that our method achieves a substantial improvement compared with other well-know deep learning methods. Beside these, we discussed two major limitations in convolutional neural networks-based semantic segmentation methods for LV segmentation.     

An Efficient Supervised Deep Hashing Method for Image Retrieval

In recent years, searching and retrieving relevant images from large databases has become an emerging challenge for the researcher. Hashing methods that mapped raw data into a short binary code have attracted increasing attention from the researcher. Most existing hashing approaches map samples to a binary vector via a single linear projection, which restricts the flexibility of those methods and leads to optimization problems. We introduce a CNN-based hashing method that uses multiple nonlinear projections to produce additional short-bit binary code to tackle this issue. Further, an end-to-end hashing system is accomplished using a convolutional neural network. Also, we design a loss function that aims to maintain the similarity between images and minimize the quantization error by providing a uniform distribution of the hash bits to illustrate the proposed technique’s effectiveness and significance. Extensive experiments conducted on various datasets demonstrate the superiority of the proposed method in comparison with state-of-the-art deep hashing methods.     

Evaluating the Learning Procedure of CNNs through a Sequence of Prognostic Tests Utilising Information Theoretical Measures

Deep learning has proven to be an important element of modern data processing technology, which has found its application in many areas such as multimodal sensor data processing and understanding, data generation and anomaly detection. While the use of deep learning is booming in many real-world tasks, the internal processes of how it draws results is still uncertain. Understanding the data processing pathways within a deep neural network is important for transparency and better resource utilisation. In this paper, a method utilising information theoretic measures is used to reveal the typical learning patterns of convolutional neural networks, which are commonly used for image processing tasks. For this purpose, training samples, true labels and estimated labels are considered to be random variables. The mutual information and conditional entropy between these variables are then studied using information theoretical measures. This paper shows that more convolutional layers in the network improve its learning and unnecessarily higher numbers of convolutional layers do not improve the learning any further. The number of convolutional layers that need to be added to a neural network to gain the desired learning level can be determined with the help of theoretic information quantities including entropy, inequality and mutual information among the inputs to the network. The kernel size of convolutional layers only affects the learning speed of the network. This study also shows that where the dropout layer is applied to has no significant effects on the learning of networks with a lower dropout rate, and it is better placed immediately after the last convolutional layer with higher dropout rates.     

Dual-Domain Fusion Convolutional Neural Network for Contrast Enhancement Forensics

Contrast enhancement forensics techniques have always been of great interest for the image forensics community, as they can be an effective tool for recovering image history and identifying tampered images. Although several contrast enhancement forensic algorithms have been proposed, their accuracy and robustness against some kinds of processing are still unsatisfactory. In order to attenuate such deficiency, in this paper, we propose a new framework based on dual-domain fusion convolutional neural network to fuse the features of pixel and histogram domains for contrast enhancement forensics. Specifically, we first present a pixel-domain convolutional neural network to automatically capture the patterns of contrast-enhanced images in the pixel domain. Then, we present a histogram-domain convolutional neural network to extract the features in the histogram domain. The feature representations of pixel and histogram domains are fused and fed into two fully connected layers for the classification of contrast-enhanced images. Experimental results show that the proposed method achieves better performance and is robust against pre-JPEG compression and antiforensics attacks, obtaining over 99% detection accuracy for JPEG-compressed images with different QFs and antiforensics attack. In addition, a strategy for performance improvements of CNN-based forensics is explored, which could provide guidance for the design of CNN-based forensics tools.     

宽带电磁图像卷积神经网络盲恢复方法研究

在利用抛物反射面对电磁干扰源成像过程中,由于系统衍射受限及成像频带较宽,导致干扰源成像模糊,分辨率低,难以分辨,不同频率不同区域干扰源所成图像分辨率不同,采用已有超分辨算法难以提高分辨率。为了实现宽带电磁图像的盲复原, 应用卷积神经网络的方法。网络训练是直接输入模糊图像,不假设任何特定的模糊和噪声模型情况下,重建出高质量图像。实验和仿真结果证明了卷积神经网络盲恢复方法在宽频带不同成像区域下表现了优于其他盲恢复算法的优势。     

Fast single image super-resolution using estimated low-frequency k-space data in MRI

PurposeSingle image super-resolution (SR) is highly desired in many fields but obtaining it is often technically limited in practice. The purpose of this study was to propose a simple, rapid and robust single image SR method in magnetic resonance (MR) imaging (MRI).MethodsThe idea is based on the mathematical formulation of the intrinsic link in k-space between a given (modulus) low-resolution (LR) image and the desired SR image. The method consists of two steps: 1) estimating the low-frequency k-space data of the desired SR image from a single LR image; 2) reconstructing the SR image using the estimated low-frequency and zero-filled high-frequency k-space data. The method was evaluated on digital phantom images, physical phantom MR images and real brain MR images, and compared with existing SR methods.ResultsThe proposed SR method exhibited a good robustness by reaching a clearly higher PSNR (25.77dB) and SSIM (0.991) averaged over different noise levels in comparison with existing edge-guided nonlinear interpolation (EGNI) (PSNR=23.78dB, SSIM=0.983), zero-filling (ZF) (PSNR=24.09dB, SSIM=0.985) and total variation (TV) (PSNR=24.54dB, SSIM=0.987) methods while presenting the same order of computation time as the ZF method but being much faster than the EGNI or TV method. The average PSNR or SSIM over different slice images of the proposed method (PSNR=26.33 dB or SSIM=0.955) was also higher than the EGNI (PSNR=25.07dB or SSIM=0.952), ZF (PSNR=24.97dB or SSIM=0.950) and TV (PSNR=25.70dB or SSIM=0.953) methods, demonstrating its good robustness to variation in anatomical structure of the images. Meanwhile, the proposed method always produced less ringing artifacts than the ZF method, gave a clearer image than the EGNI method, and did not exhibit any blocking effect presented in the TV method. In addition, the proposed method yielded the highest spatial consistency in the inter-slice dimension among the four methods.ConclusionsThis study proposed a fast, robust and efficient single image SR method with high spatial consistency in the inter-slice dimension for clinical MR images by estimating the low-frequency k-space data of the desired SR image from a single spatial modulus LR image.     

Sensor and Actuator Fault Diagnosis for Robot Joint Based on Deep CNN

This paper proposes a data-driven method-based fault diagnosis method using the deep convolutional neural network (DCNN). The DCNN is used to deal with sensor and actuator faults of robot joints, such as gain error, offset error, and malfunction for both sensors and actuators, and different fault types are diagnosed using the trained neural network. In order to achieve the above goal, the fused data of sensors and actuators are used, where both types of fault are described in one formulation. Then, the deep convolutional neural network is applied to learn characteristic features from the merged data to try to find discriminative information for each kind of fault. After that, the fully connected layer does prediction work based on learned features. In order to verify the effectiveness of the proposed deep convolutional neural network model, different fault diagnosis methods including support vector machine (SVM), artificial neural network (ANN), conventional neural network (CNN) using the LeNet-5 method, and long-term memory network (LTMN) are investigated and compared with DCNN method. The results show that the DCNN fault diagnosis method can realize high fault recognition accuracy while needing less model training time.     

基于全卷积金字塔残差网络的能谱CT图像降噪研究

传统CT采用积分式探测器采集投影数据,反映的是物体的平均衰减特性,会在一定程度上造成信息损失,无法对物体进行较好的定性定量测量。基于光子计数探测器的能谱CT通过设定多个能量响应阈值能够探测不同能量范围内的X射线光子,采集更多被测物体的物质组成信息,有助于识别不同物理特性的材料,基于此,能谱CT被广泛的应用于小病灶、低对比度结构以及微细结构的成像。然而将整个能谱划分为多个能量段进行数据采集时,范围较窄能量范围内的有效光子数比例相对降低,导致图像中包含较多的噪声,图像质量较差,影响能谱CT的临床应用。为了有效的抑制能谱CT不同能量段内重建图像中的噪声,提出了一种基于深度学习的能谱CT图像降噪方法。我们将全卷积网络和金字塔残差网络结合为全卷积金字塔残差网络(FCPRN),实验中,利用能谱CT在不同的能量范围扫描小鼠样本,使用FDK算法和基于压缩感知的Split-Bregman算法进行重建并分别作为训练数据和标签数据训练全卷积金字塔残差网络。为了验证网络的降噪性能,选取了常见的降噪网络模型denoising convolutional neural networks(DNCN)以及residual encoder decoder convolutional neural network (REDCNN)进行对比,训练三种网络的使用的数据和实验配置都是完全相同的,实验结果表明训练模型可以有效抑制不同能量范围内重建图像的噪声,且使用的全卷积金字塔残差网络的降噪性能优于其他网络模型。模型训练好后,可以对FDK算法重建出的图像进行降噪,由此提高能谱CT图像降噪效率,保证能谱CT重建图像的质量。     

一种基于2D-CNN的激光超声表面缺陷检测方法

激光超声表面缺陷检测的过程中,缺陷的定量表征通常依赖于操作者的判断,易受到人为因素干扰,致使检测结果不稳定.针对这一问题,提出一种基于图像识别的二维卷积神经网络(2D-CNN)的缺陷自动分类检测方法.利用有限元方法模拟激光超声检测过程,并采集超声信号数据用于训练分类模型;使用连续小变换(CWT)处理超声信号得到小波时频...     

Deep-learning-based prediction of living cells mitosis via quantitative phase microscopy

We present a deep learning approach for living cells mitosis classification based on label-free quantitative phase imaging with transport of intensity equation methods. In the approach, we applied a pretrained deep convolutional neural network using transfer learning for binary classification of mitosis and non-mitosis. As a validation, we demonstrated the performances of the network trained by phase images and intensity images, respectively. The convolutional neural network trained by phase images achieved an average accuracy of 98.9% on the validation data, which outperforms the average accuracy 89.6% obtained by the network trained by intensity images. We believe that the quantitative phase microscopy in combination with deep learning enables researchers to predict the mitotic status of living cells noninvasively and efficiently.