深度学习在超声检测缺陷识别中的应用与发展*

深度学习（Deep Learning）是目前最强大的机器学习算法之一,其中卷积神经网络（Convolutional Neural Network, CNN）模型具有自动学习特征的能力,在图像处理领域较其他深度学习模型有较大的性能优势。本文先简述了深度学习的发展史,然后综述了深度学习在超声检测缺陷识别中的应用与发展,从早期浅层神经网络到现在深度学习的应用现状,并借鉴医学影像识别和射线图像识别领域的方法,分析了卷积神经网络对超声图像缺陷识别的适用性。最后,探讨归纳了目前在超声检测图像识别中使用CNN存在的一些问题,及其主要应对策略的研究方向。     

卷积神经网络用于近红外光谱预测土壤含水率

近红外光谱分析技术在土壤含水率预测方面具有独特的优势,是一种便捷且有效的方法。卷积神经网络作为高性能的深度学习模型,能够从复杂光谱数据中自主提取有效特征结构进行学习,与传统的浅层学习模型相比具有更强的模型表达能力。将卷积神经网络用于近红外光谱预测土壤含水率,并提出了有效的卷积神经网络光谱回归建模方法,简化了光谱数据的预处理要求,且具有更高的光谱预测精度。首先对不同含水率下土壤样品的光谱反射率数据进行简单的预处理,通过主成分分析减少光谱数据量,并将处理后的光谱数据变换为二维光谱信息矩阵,以适应卷积神经网络特殊的学习结构。然后基于卷积神经网络算法,设置双层卷积和池化结构逐层提取光谱数据的内部特征信息,并采用局部连接和权值共享减少网络参数、提高泛化性能。通过试验优化网络结构和各项参数,最终获得针对土壤光谱数据的卷积神经网络土壤含水率预测模型,并与传统的BP,PLSR和LSSVM模型进行对比实验。结果表明在训练样本达到一定数量时,卷积神经网络的预测精度和回归拟合度均高于三种传统模型。在少量训练样本参与建模的情况下,模型预测表现高于BP神经网络,但略低于PLSR和LSSVM模型。随着参与训练样本量的增加,卷积神经网络的预测精度和回归拟合度也随之稳定提升,达到并显著优于传统模型水平。因此,卷积神经网络能够利用近红外光谱数据对土壤含水率做出有效预测,且在较多样本参与建模时取得更好效果。     

Permutation Entropy-Based Interpretability of Convolutional Neural Network Models for Interictal EEG Discrimination of Subjects with Epileptic Seizures vs. Psychogenic Non-Epileptic Seizures

The differential diagnosis of epileptic seizures (ES) and psychogenic non-epileptic seizures (PNES) may be difficult, due to the lack of distinctive clinical features. The interictal electroencephalographic (EEG) signal may also be normal in patients with ES. Innovative diagnostic tools that exploit non-linear EEG analysis and deep learning (DL) could provide important support to physicians for clinical diagnosis. In this work, 18 patients with new-onset ES (12 males, 6 females) and 18 patients with video-recorded PNES (2 males, 16 females) with normal interictal EEG at visual inspection were enrolled. None of them was taking psychotropic drugs. A convolutional neural network (CNN) scheme using DL classification was designed to classify the two categories of subjects (ES vs. PNES). The proposed architecture performs an EEG time-frequency transformation and a classification step with a CNN. The CNN was able to classify the EEG recordings of subjects with ES vs. subjects with PNES with 94.4% accuracy. CNN provided high performance in the assigned binary classification when compared to standard learning algorithms (multi-layer perceptron, support vector machine, linear discriminant analysis and quadratic discriminant analysis). In order to interpret how the CNN achieved this performance, information theoretical analysis was carried out. Specifically, the permutation entropy (PE) of the feature maps was evaluated and compared in the two classes. The achieved results, although preliminary, encourage the use of these innovative techniques to support neurologists in early diagnoses.     

Efficient Network Architecture Search Using Hybrid Optimizer

Manually designing a convolutional neural network (CNN) is an important deep learning method for solving the problem of image classification. However, most of the existing CNN structure designs consume a significant amount of time and computing resources. Over the years, the demand for neural architecture search (NAS) methods has been on the rise. Therefore, we propose a novel deep architecture generation model based on Aquila optimization (AO) and a genetic algorithm (GA). The main contributions of this paper are as follows: Firstly, a new encoding strategy representing the CNN coding structure is proposed, so that the evolutionary computing algorithm can be combined with CNN. Secondly, a new mechanism for updating location is proposed, which incorporates three typical operators from GA cleverly into the model we have designed so that the model can find the optimal solution in the limited search space. Thirdly, the proposed method can deal with the variable-length CNN structure by adding skip connections. Fourthly, combining traditional CNN layers and residual blocks and introducing a grouping strategy provides greater possibilities for searching for the optimal CNN structure. Additionally, we use two notable datasets, consisting of the MNIST and CIFAR-10 datasets for model evaluation. The experimental results show that our proposed model has good results in terms of search accuracy and time.     

基于GADF变换和多尺度CNN的哈密瓜表面农药残留可见-近红外光谱判别方法

针对哈密瓜表面农药残留化学检测方法成本高且具有破坏性等问题,探索了可见-近红外（Vis-NIR）光谱技术对农药残留定性判别的可行性。以哈密瓜为载体,百菌清和吡虫啉农药为研究对象,采集哈密瓜表面无残留、百菌清和吡虫啉残留的可见-近红外漫反射光谱,利用格拉姆角场（GAF）将一维光谱数据转换为二维彩色图像,构建GAF图像数据集。设计一种包含Inception结构的多尺度卷积神经网络模型用于哈密瓜表面农药残留种类判别,包括1层输入层、3层卷积层、1层融合层、1层平坦层、2层全连接层和1层输出层。模型测试混淆矩阵结果表明,格拉姆角差场（GADF）变换对哈密瓜表面农药残留的可见-近红外光谱表达能力较强。此外,构建AlexNet、VGG-16卷积神经网络（CNN）模型和支持向量机（SVM）、极限学习机（ELM）机器学习模型与提出的多尺度CNN模型进行性能对比。结果表明,3种CNN模型对哈密瓜表面有无农药残留的判别效果较好,综合判别准确率均高于SVM和ELM模型。对比3种CNN模型性能,多尺度CNN模型的性能最佳,训练耗时为14 s,综合判别准确率为98.33%。多尺度CNN模型结构利用多种小尺寸滤波器组合（1×1,3×3和5×5）和并行卷积模块,能够捕获不同层次和尺度的特征,通过级联融合模式进行深度特征融合,提高了模型的特征提取能力。与传统深度CNN模型相比,在保证计算复杂度不变的情况下,多尺度CNN模型的精度得到了有效提高。实验结果表明,GADF变换结合多尺度CNN模型可以有效进行光谱数据解析,利用可见-近红外光谱技术可以实现哈密瓜表面农药残留的定性判别。研究结果为大型瓜果表面农药残留的快速无损检测技术的研发提供了理论参考。     

基于深度学习的太赫兹时域光谱识别研究

太赫兹时域光谱技术,由于其具有物质"指纹谱"特性,是一种可以快速无损地鉴别物质的重要手段,在毒品和爆炸物的无损检测等方面有广阔的应用前景.其中,光谱识别是太赫兹时域光谱技术应用研究的重要方向之一.现有的光谱识别方法多是依靠手工选取特征后进行机器学习分类,或是通过设置吸收峰阈值门限进行判断.由于一些物质在太赫兹波段内并没...     

Widrow-cellular neural network and optoelectronic implementation

A new type of optoelectronic cellular neural network has been developed by providing the capability of coefficients adjusment of cellular neural network (CNN) using Widrow based perceptron learning algorithm. The new supervised cellular neural network is called Widrow-CNN. Despite the unsupervised CNN, the proposed learning algorithm allows to use the Widrow-CNN for various image processing applications easily. Also, the capability of CNN for image processing and feature extraction has been improved using basic joint transform correlation architecture. This hardware application presents high speed processing capability compared to digital applications. The optoelectronic Widrow-CNN has been tested for classic CNN feature extraction problems. It yields the best results even in case of hard feature extraction problems such as diagonal line detection and vertical line determination.     

Sentiment Analysis of Persian Movie Reviews Using Deep Learning

Sentiment analysis aims to automatically classify the subject’s sentiment (e.g., positive, negative, or neutral) towards a particular aspect such as a topic, product, movie, news, etc. Deep learning has recently emerged as a powerful machine learning technique to tackle the growing demand for accurate sentiment analysis. However, the majority of research efforts are devoted to English-language only, while information of great importance is also available in other languages. This paper presents a novel, context-aware, deep-learning-driven, Persian sentiment analysis approach. Specifically, the proposed deep-learning-driven automated feature-engineering approach classifies Persian movie reviews as having positive or negative sentiments. Two deep learning algorithms, convolutional neural networks (CNN) and long-short-term memory (LSTM), are applied and compared with our previously proposed manual-feature-engineering-driven, SVM-based approach. Simulation results demonstrate that LSTM obtained a better performance as compared to multilayer perceptron (MLP), autoencoder, support vector machine (SVM), logistic regression and CNN algorithms.     

Hybrid quantum-classical convolutional neural networks

Deep learning has been shown to be able to recognize data patterns better than humans in specific circumstances or contexts. In parallel, quantum computing has demonstrated to be able to output complex wave functions with a few number of gate operations,which could generate distributions that are hard for a classical computer to produce. Here we propose a hybrid quantum-classical convolutional neural network(QCCNN), inspired by convolutional neural networks(CNNs) but adapted to quantum computing to enhance the feature mapping process. QCCNN is friendly to currently noisy intermediate-scale quantum computers, in terms of both number of qubits as well as circuit's depths, while retaining important features of classical CNN, such as nonlinearity and scalability. We also present a framework to automatically compute the gradients of hybrid quantum-classical loss functions which could be directly applied to other hybrid quantum-classical algorithms. We demonstrate the potential of this architecture by applying it to a Tetris dataset, and show that QCCNN can accomplish classification tasks with learning accuracy surpassing that of classical CNN with the same structure.     

Binary neutron stars gravitational wave detection based on wavelet packet analysis and convolutional neural networks

This work investigates the detection of binary neutron stars gravitational wave based on convolutional neural network(CNN).To promote the detection performance and efficiency,we proposed a scheme based on wavelet packet(WP)decomposition and CNN.The WP decomposition is a time-frequency method and can enhance the discriminant features between gravitational wave signal and noise before detection.The CNN conducts the gravitational wave detection by learning a function mapping relation from the data under being processed to the space of detection results.This function-mapping-relation style detection scheme can detection efficiency significantly.In this work,instrument effects are con-sidered,and the noise are computed from a power spectral density(PSD)equivalent to the Advanced LIGO design sensitivity.The quantitative evaluations and comparisons with the state-of-art method matched filtering show the excellent performances for BNS gravitational wave detection.On efficiency,the current experiments show that this WP-CNN-based scheme is more than 960 times faster than the matched filtering.     

基于级联卷积神经网络的前列腺磁共振图像分类

针对深度学习训练成本高,以及基于磁共振图像的前列腺癌临床诊断需要大量医学常识且极为耗时的问题,本文提出了一种基于级联卷积神经网络（Convolutional Neural Network,CNN）和磁共振图像的前列腺癌（Prostate Cancer,PCa）自动分类诊断方法,该网络以Faster-RCNN作为前网络,对前列腺区域进行提取分割,用于排除前列腺附近组织器官的干扰;以基于ResNet改进的网络结构CNN40_bottleneck作为后网络,用于对前列腺区域病变进行分类.后网络由瓶颈结构串联组成,其中使用批量标准化（Batch Normalization,BN）、全局平均池化（Global Average Pooling,GAP）进行优化.实验结果证明,本文方法对前列腺癌诊断结果较好,而且缩减了训练时间和参数量,有效降低了训练成本.     

ML-Based Analysis of Particle Distributions in High-Intensity Laser Experiments: Role of Binning Strategy

When entering the phase of big data processing and statistical inferences in experimental physics, the efficient use of machine learning methods may require optimal data preprocessing methods and, in particular, optimal balance between details and noise. In experimental studies of strong-field quantum electrodynamics with intense lasers, this balance concerns data binning for the observed distributions of particles and photons. Here we analyze the aspect of binning with respect to different machine learning methods (Support Vector Machine (SVM), Gradient Boosting Trees (GBT), Fully-Connected Neural Network (FCNN), Convolutional Neural Network (CNN)) using numerical simulations that mimic expected properties of upcoming experiments. We see that binning can crucially affect the performance of SVM and GBT, and, to a less extent, FCNN and CNN. This can be interpreted as the latter methods being able to effectively learn the optimal binning, discarding unnecessary information. Nevertheless, given limited training sets, the results indicate that the efficiency can be increased by optimizing the binning scale along with other hyperparameters. We present specific measurements of accuracy that can be useful for planning of experiments in the specified research area.     

基于卷积神经网络与光谱特征的夏威夷果品质鉴定研究

夏威夷果含油量高,在开缝之后容易发生变质,现有关于夏威夷果品质鉴定的方法多为传统的破坏性检验,很难满足无损检测的需求。卷积神经网络(CNN)作为应用最广泛的深度学习网络模型之一,具有比浅层学习方法更强的特征提取与模型表达能力,在光谱数据方面的应用拥有很大潜力。基于夏威夷果在可见-近红外的光谱特征分析,研究用于提取夏威夷果光谱特征的卷积神经网络模型,并提出一种高效无损鉴定夏威夷果品质的方法。首先以三种不同品质的夏威夷果(好籽、哈籽及霉籽)为研究对象,分析样本在500~2 100 nm的光谱信息;在光谱数据预处理中引入白化处理方法,用以增强数据的相关性差异;然后在模型训练过程中,将样本随机分为训练集和预测集,探讨不同CNN结构、卷积层数、卷积核大小及个数、池化层类型、全连接层神经元个数以及激活函数对分类结果的影响,并采用激活函数ReLU和Dropout方法,预防样本数据过少引起的过拟合现象;最后通过分析模型分类准确率和计算效率,确定了一个6层结构的CNN模型： 输入层-卷积层-池化层-全连接层(200神经元)-全连接层(100神经元)-输出层。实验结果表明： 上述网络模型对校正集和预测集的分类准确率均达到100％。因此,改进后的卷积神经网络模型可充分学习夏威夷果的光谱特征并有效分类,将深度学习理论与光谱分析相结合的方法能够实现对夏威夷果品质的准确鉴定,同时为夏威夷果等坚果类食品的高效、无损、实时在线检测提供了新思路。     

High-performance rapid MR parameter mapping using model-based deep adversarial learning

PurposeTo develop and evaluate a deep adversarial learning-based image reconstruction approach for rapid and efficient MR parameter mapping.MethodsThe proposed method provides an image reconstruction framework by combining the end-to-end convolutional neural network (CNN) mapping, adversarial learning, and MR physical models. The CNN performs direct image-to-parameter mapping by transforming a series of undersampled images directly into MR parameter maps. Adversarial learning is used to improve image sharpness and enable better texture restoration during the image-to-parameter conversion. An additional pathway concerning the MR signal model is added between the estimated parameter maps and undersampled k-space data to ensure the data consistency during network training. The proposed framework was evaluated on T₂ mapping of the brain and the knee at an acceleration rate R = 8 and was compared with other state-of-the-art reconstruction methods. Global and regional quantitative assessments were performed to demonstrate the reconstruction performance of the proposed method.ResultsThe proposed adversarial learning approach achieved accurate T₂ mapping up to R = 8 in brain and knee joint image datasets. Compared to conventional reconstruction approaches that exploit image sparsity and low-rankness, the proposed method yielded lower errors and higher similarity to the reference and better image sharpness in the T₂ estimation. The quantitative metrics were normalized root mean square error of 3.6% for brain and 7.3% for knee, structural similarity index of 85.1% for brain and 83.2% for knee, and tenengrad measures of 9.2% for brain and 10.1% for the knee. The adversarial approach also achieved better performance for maintaining greater image texture and sharpness in comparison to the CNN approach without adversarial learning.ConclusionThe proposed framework by incorporating the efficient end-to-end CNN mapping, adversarial learning, and physical model enforced data consistency is a promising approach for rapid and efficient reconstruction of quantitative MR parameters.     

基于DCGAN的拉曼光谱样本扩充及应用研究

拉曼光谱检测方法依赖于化学计量学算法,深度学习是当下最炙手可热的方向,可应用于拉曼光谱进行建模.但是深度学习需要大样本进行训练,而拉曼光谱采集受制于器材和人力成本,获取大批量的样本需要更大成本,且易受荧光等因素干扰,这些问题都制约了将深度学习应用于拉曼光谱.针对以上问题,通过引入深度卷积生成对抗网络(DCGAN)提取拉...     

Determination of quantum toric error correction code threshold using convolutional neural network decoders

Quantum error correction technology is an important solution to solve the noise interference generated during the operation of quantum computers.In order to find the best syndrome of the stabilizer code in quantum error correction,we need to find a fast and close to the optimal threshold decoder.In this work,we build a convolutional neural network(CNN)decoder to correct errors in the toric code based on the system research of machine learning.We analyze and optimize various conditions that affect CNN,and use the RestNet network architecture to reduce the running time.It is shortened by 30%-40%,and we finally design an optimized algorithm for CNN decoder.In this way,the threshold accuracy of the neural network decoder is made to reach 10.8%,which is closer to the optimal threshold of about 11%.The previous threshold of 8.9%-10.3%has been slightly improved,and there is no need to verify the basic noise.     

A convolutional neural network approach to calibrating the rotation axis for X‐ray computed tomography

This paper presents an algorithm to calibrate the center‐of‐rotation for X‐ray tomography by using a machine learning approach, the Convolutional Neural Network (CNN). The algorithm shows excellent accuracy from the evaluation of synthetic data with various noise ratios. It is further validated with experimental data of four different shale samples measured at the Advanced Photon Source and at the Swiss Light Source. The results are as good as those determined by visual inspection and show better robustness than conventional methods. CNN has also great potential for reducing or removing other artifacts caused by instrument instability, detector non‐linearity, etc. An open‐source toolbox, which integrates the CNN methods described in this paper, is freely available through GitHub at tomography/xlearn and can be easily integrated into existing computational pipelines available at various synchrotron facilities. Source code, documentation and information on how to contribute are also provided.     

结合高光谱成像和机器学习的棉种年份鉴别

棉花精量播种技术目前已经在新疆兵团全面推广,该技术能精确实现一穴一粒的农艺技术指标,但是也对高质量棉种的筛选提出了更高的要求。为了避免播种往年活力不足的棉种而导致发芽率降低的问题,结合机器学习和近红外（NIR）高光谱成像技术（HSI）进行棉种年份精确鉴别,实现棉种的快速无损筛选。采集2016年—2019年近四年外观无明显差异的棉种各360粒,共1 440粒棉种（按照3∶1∶1划分训练集、验证集和测试集）作为样本,按照每批60粒采集915~1 698 nm范围的棉种高光谱图像,去除首尾两端噪声大的光谱,保留1 002~1 602 nm范围的光谱为原始数据。利用Savitzky-Golay（SG）平滑算法对光谱进行预处理,采用主成分载荷方法（PCA-loading）选取13个特征波段,基于全部光谱数据和特征波段（±10 nm）数据建立逻辑回归（LR）、偏最小二乘判别分析（PLS-DA）、支持向量机（SVM）、循环神经网络（RNN）、长短记忆网络（LSTM）和卷积神经网络（CNN）六种分类模型。使用全光谱数据建模时,六种分类模型在测试集上的鉴别准确率分别为96.27%,98.98%,99.32%,96.95%,97.63%和100%,其中CNN和SVM模型取得了较好的结果;使用特征光谱数据建模时,六种分类模型在测试集上的鉴别精度分别为93.56%,97.29%,98.30%,95.25%,94.24%和99.66%,其中CNN和SVM模型仍有较好的分类结果。结果表明,使用全光谱数据建模时,六种分类模型都可以实现较高精度的棉种年份鉴别,使用特征光谱数据建模时CNN和SVM模型的鉴别精度仍可达到98%;其中深度学习方法优于传统机器学习方法,但是传统机器学习方法仍能保持较好的鉴别准确率。因此,结合近红外高光谱成像技术和机器学习方法能够实现棉种年份的高精度鉴别,为棉花精量播种过程中的优质棉种选种技术提供理论依据和方法。     

Deep-learning approach with convolutional neural network for classification of maximum intensity projections of dynamic contrast-enhanced breast magnetic resonance imaging

PurposeWe aimed to evaluate deep learning approach with convolutional neural networks (CNNs) to discriminate between benign and malignant lesions on maximum intensity projections of dynamic contrast-enhanced breast magnetic resonance imaging (MRI).MethodsWe retrospectively gathered maximum intensity projections of dynamic contrast-enhanced breast MRI of 106 benign (including 22 normal) and 180 malignant cases for training and validation data. CNN models were constructed to calculate the probability of malignancy using CNN architectures (DenseNet121, DenseNet169, InceptionResNetV2, InceptionV3, NasNetMobile, and Xception) with 500 epochs and analyzed that of 25 benign (including 12 normal) and 47 malignant cases for test data. Two human readers also interpreted these test data and scored the probability of malignancy for each case using Breast Imaging Reporting and Data System. Sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve (AUC) were calculated.ResultsThe CNN models showed a mean AUC of 0.830 (range, 0.750–0.895). The best model was InceptionResNetV2. This model, Reader 1, and Reader 2 had sensitivities of 74.5%, 72.3%, and 78.7%; specificities of 96.0%, 88.0%, and 80.0%; and AUCs of 0.895, 0.823, and 0.849, respectively. No significant difference arose between the CNN models and human readers (p > 0.125).ConclusionOur CNN models showed comparable diagnostic performance in differentiating between benign and malignant lesions to human readers on maximum intensity projection of dynamic contrast-enhanced breast MRI.     

Learning in Convolutional Neural Networks Accelerated by Transfer Entropy

Recently, there is a growing interest in applying Transfer Entropy (TE) in quantifying the effective connectivity between artificial neurons. In a feedforward network, the TE can be used to quantify the relationships between neuron output pairs located in different layers. Our focus is on how to include the TE in the learning mechanisms of a Convolutional Neural Network (CNN) architecture. We introduce a novel training mechanism for CNN architectures which integrates the TE feedback connections. Adding the TE feedback parameter accelerates the training process, as fewer epochs are needed. On the flip side, it adds computational overhead to each epoch. According to our experiments on CNN classifiers, to achieve a reasonable computational overhead–accuracy trade-off, it is efficient to consider only the inter-neural information transfer of the neuron pairs between the last two fully connected layers. The TE acts as a smoothing factor, generating stability and becoming active only periodically, not after processing each input sample. Therefore, we can consider the TE is in our model a slowly changing meta-parameter.