改进YOLOv4的铁路沿线遥感影像地物检测方法

近年来,高分遥感影像技术的快速发展为铁路沿线地物检测提供了一种重要技术手段。基于回归的一阶段目标检测方法YOLOv4具有检测精度高、速度快等优点,但用于遥感影像检测时仍然存在部分细节特征信息丢失导致的小目标漏检,以及进行大面积地物检测时效率低的问题。为此,提出改进YOLOv4网络模型对遥感影像铁路沿线地物进行检测。首先,设计由卷积、批量归一化和Mish激活函数组成的CBM(convolution batch normalization mish)模块,并采用DCBM(double CBM)模块作为密集连接网络(DenseNet)的传输层用于YOLOv4网络特征提取以实现地物特征传递和信息重用,增强小目标地物的检测能力,降低漏检率;然后针对YOLOv4在大面积检测时效率不高和模型参数空间较大的缺陷,将压缩激励SE(squeeze excitation)通道注意机制用于骨干网中跨阶段局部单元(cross stage partial, CSP)的每个残差单元之后,减少SE注意模块的重复调用次数,使其能够在提高网络性能的同时降低模型参数量从而提高检测效率;最后,针对长条形状的铁路目标提取困难问题,在网络结果输出之前引入改进的通道空间注意力机制ICBAM(improved convolutional block attention module) 保留原始特征信息,解决铁路目标特征提取能力差的问题,提高铁路中大尺度目标的检测效率。为验证所提方法的有效性,选取2 048张分辨率为1 920×1 080的某段铁路沿线遥感影像地物样本数据,将其中的铁路、房屋、楼宇建筑、农田和水池作为检测目标进行实验,并与当前流行的目标检测方法进行对比。结果表明,改进方法不仅增强了对小目标地物的检测能力,提高了地物检测精度和速度,而且提高了大尺度目标的检测效率。与YOLOv4算法相比,mAP提高了2.11%,准确率提高了2.93%,召回率提高了3.79%,模型大小减少了8.53%。所提方法为当前应用高速铁路沿线遥感影像地物快速精准检测提供了有效方法。     

A Study on Railway Surface Defects Detection Based on Machine Vision

The detection of rail surface defects is an important tool to ensure the safe operation of rail transit. Due to the complex diversity of track surface defect features and the small size of the defect area, it is difficult to obtain satisfying detection results by traditional machine vision methods. The existing deep learning-based methods have the problems of large model sizes, excessive parameters, low accuracy and slow speed. Therefore, this paper proposes a new method based on an improved YOLOv4 (You Only Look Once, YOLO) for railway surface defect detection. In this method, MobileNetv3 is used as the backbone network of YOLOv4 to extract image features, and at the same time, deep separable convolution is applied on the PANet layer in YOLOv4, which realizes the lightweight network and real-time detection of the railway surface. The test results show that, compared with YOLOv4, the study can reduce the amount of the parameters by 78.04%, speed up the detection by 10.36 frames per second and decrease the model volume by 78%. Compared with other methods, the proposed method can achieve a higher detection accuracy, making it suitable for the fast and accurate detection of railway surface defects.     

Robust Vehicle Speed Measurement Based on Feature Information Fusion for Vehicle Multi-Characteristic Detection

A robust vehicle speed measurement system based on feature information fusion for vehicle multi-characteristic detection is proposed in this paper. A vehicle multi-characteristic dataset is constructed. With this dataset, seven CNN-based modern object detection algorithms are trained for vehicle multi-characteristic detection. The FPN-based YOLOv4 is selected as the best vehicle multi-characteristic detection algorithm, which applies feature information fusion of different scales with both rich high-level semantic information and detailed low-level location information. The YOLOv4 algorithm is improved by combing with the attention mechanism, in which the residual module in YOLOv4 is replaced by the ECA channel attention module with cross channel interaction. An improved ECA-YOLOv4 object detection algorithm based on both feature information fusion and cross channel interaction is proposed, which improves the performance of YOLOv4 for vehicle multi-characteristic detection and reduces the model parameter size and FLOPs as well. A multi-characteristic fused speed measurement system based on license plate, logo, and light is designed accordingly. The system performance is verified by experiments. The experimental results show that the speed measurement error rate of the proposed system meets the requirement of the China national standard GB/T 21555-2007 in which the speed measurement error rate should be less than 6%. The proposed system can efficiently enhance the vehicle speed measurement accuracy and effectively improve the vehicle speed measurement robustness.     

面向嵌入式平台的轻量化神经网络手势识别方法

针对传统基于图像分割和特征提取的手势识别算法在复杂背景下识别准确率低、灵活性差的问题,基于目标检测神经网络的手势识别算法可以有效提高复杂环境下手势识别的准确性。受嵌入式处理器体积和功耗的限制,常用的目标检测神经网络在嵌入式上的识别速度较低,不能满足实时手势识别的要求。在SSD目标检测的基础上对其进行优化,使用MobileNetv3网络实现特征提取,目标检测方面则是使用SSD-lite结构,其使用深度可分离卷积替代普通卷积,实现了轻量化MobileNetv3-SSDLite手势识别算法的设计。针对手势识别的要求,制作了包含不同手势的数据集,利用它在服务器上完成了模型的训练。为了满足嵌入式的算力限制,通过模型的量化压缩将float64的网络参数量化为int8,并压缩网络结构,提高网络在嵌入式上的推理速度,实现基于嵌入式的手势识别。实验结果表明,基于嵌入式的MobileNetv3-SSDLite手势识别算法可以达到平均准确率99.61%,且识别速度达到每秒50帧以上,满足实时手势识别的要求。     

Learning Visible Thermal Person Re-Identification via Spatial Dependence and Dual-Constraint Loss

Visible thermal person re-identification (VT Re-ID) is the task of matching pedestrian images collected by thermal and visible light cameras. The two main challenges presented by VT Re-ID are the intra-class variation between pedestrian images and the cross-modality difference between visible and thermal images. Existing works have principally focused on local representation through cross-modality feature distribution, but ignore the internal connection of the local features of pedestrian body parts. Therefore, this paper proposes a dual-path attention network model to establish the spatial dependency relationship between the local features of the pedestrian feature map and to effectively enhance the feature extraction. Meanwhile, we propose cross-modality dual-constraint loss, which adds the center and boundary constraints for each class distribution in the embedding space to promote compactness within the class and enhance the separability between classes. Our experimental results show that our proposed approach has advantages over the state-of-the-art methods on the two public datasets SYSU-MM01 and RegDB. The result for the SYSU-MM01 is Rank-1/mAP 57.74%/54.35%, and the result for the RegDB is Rank-1/mAP 76.07%/69.43%.     

改进YOLO轻量化网络的行人检测算法

针对当前行人检测方法计算量大、检测精度低的问题,基于YOLOv4-tiny提出一种改进的行人检测算法.引入通道注意力和空间注意力模块(CBAM)至CSPDarknet53-tiny网络中,通过学习图像的位置信息和通道信息得到更加丰富的特征;在骨干网络CSPDarknet53-tiny之后引入空间金字塔池化模块,能够极大...     

结合注意力机制的改进U-Net网络在端到端语音增强中的应用

设计了一个适用于端到端语音增强的改进的U-Net (Attention Dilated Convolution U-Net,ADC-U-Net)网络模型.与基线U-Net网络相比,一方面通过加入空洞卷积减小由采样带来的信息损失;另一方面引入了注意力机制结构,结合了含噪语音更多的上下文信息,提取更深层次和更丰富的特征信息...     

Bridge disease detection and recognition based on improved YOLOX algorithm北大核心

In view of the low accuracy of the current bridge disease detection algorithm based on convolutional neural network, an improved YOLOX algorithm was proposed to improve the detection accuracy. By using the feature information of the shallow layer of the backbone network, the feature extraction enhancement network was improved, and the feature information of the same layer was added for fusion. An improved coordinate attention mechanism was introduced to combine the position information and the channel information to enhance the network recognition of bridge diseases. At the same time, the localization loss function was improved. The experimental results show that the accuracy of the improved YOLOX network structure for bridge disease detection reaches 92.11%, which is 4.40% higher than the original network. © 2023 Editorial office of Journal of Applied Optics. All rights reserved.     

基于一维卷积神经网络的雌激素粉末拉曼光谱定性分类

拉曼光谱物质定性鉴别已被广泛应用于诸多行业和研究领域,但传统拉曼光谱分析过程中的预处理主要依赖人为经验,光谱特征提取虽然能够降低信号维度,同时也会造成部分光谱信息损失。特性相近物质本身光谱相似度较高,受到测量过程中环境干扰和分析过程中多种误差影响,导致最终分类效果并不理想。针对此问题,提出基于一维卷积神经网络（one-dimensional convolution neural network,1D-CNN）的拉曼光谱定性分类方法。实验采集雌酮（Estrone）、雌二醇（Estradiol）,雌三醇（Estriol）三种不同雌性激素粉末的拉曼光谱,设计随机平移、添加噪声和随机加权三种光谱数据增强方法,构建数量充足的拉曼光谱数据库用于神经网络模型训练与测试;基于拉曼光谱数据特点提出一维卷积神经网络分类模型,将光谱预处理、特征提取和定性分类的全过程融为一体。通过大量仿真实验,优化所提出的神经网络模型超参数和训练过程并测试分类效果,从预处理对光谱分类结果的影响和模型抗干扰性能两个方面与多种传统拉曼光谱分类算法对比,评价模型性能。实验结果表明,本文提出的一维卷积神经网络模型可实现三类雌性激素粉末拉曼光谱快速准确分类,分类正确率最高可达98.26%,分析过程中无需光谱预处理和特征提取步骤,简化了光谱分析流程,并能保留更多有效信息。同时,当模拟测量噪声强度达到60 dBW时,传统方法分类正确率均明显出现不同程度明显降低,卷积神经网络模型依然能够取得96.81%的分类正确率,说明相比对传统拉曼光谱分类方法,所提出方法受光谱测量噪声影响更小,鲁棒性更强,适用于分析更复杂现场测量的强噪声拉曼光谱信号。该研究结果表明深度学习方法在拉曼光谱的分析与处理领域具有很大的应用潜力和研究价值。     

一维空洞卷积神经网络的矿物光谱分类

矿物光谱综合反映了岩矿的物理化学特性、组分和内部结构特征,已被应用于岩矿识别研究.传统的矿物光谱分类方法需要先对矿物光谱进行预处理,再采用不同方法分析光谱特征,从而实现分类目的.但同时也会造成部分光谱信息丢失,导致最终分类精度不高且操作过程繁琐、效率低下,难以应对日益增长的大数据处理需求.因此,建立一个准确、高效的矿物...     

基于Mel频谱值和深度学习网络的鸟声识别算法*

为了增强网络对鸟鸣声信号的特征学习能力并提高识别精度,提出一种基于深度残差收缩网络和扩张卷积的鸟声识别方法。首先,提取鸟鸣声信号的对数梅尔特征及其一阶和二阶差分系数组成logMel特征集作为网络模型的输入;其次,通过深度残差收缩网络自动学习噪声阈值,减少噪声干扰;然后,引入扩张卷积增大卷积核感受野并利用注意力机制使网络更关注关键帧特征;最后,通过双向长短时记忆网络从学到的局部特征中学习长期依赖关系。以百鸟数据birdsdata鸟声库中的19种中国常见鸟类作为实验对象,识别正确率可以达到96.58%,并对比模型在不同信噪比数据下的识别结果,结果表明该模型在噪声环境下的识别效果优于现有模型。     

基于优化YOLOv4算法的行驶车辆要素检测方法北大核心CSCD

随着车辆数量的急剧增加,带来了一系列管理问题,智能交通系统是一种有效的解决方式。由于传统的目标识别方式受天气、距离、角度、光照等因素的影响较大,且基于原YOLOv4算法的驾驶员面部、手部等信息检测的准确率不高,提出一种基于优化YOLOv4算法的检测定位方法。在给原YOLOv4网络增加一个更小的检测尺度的同时,使用模糊ISODATA动态聚类算法对先验框数目进行优化,并使用真实十字路口数据集进行实验。实验证明,优化后的网络在训练集中的类间平均准确率为98.56%,检测帧频为41.43帧/s,均高于原网络。     

基于多模态MRI与深度学习的乳腺病变良恶性鉴别

为提高基于动态增强磁共振成像（DCE-MRI）的计算机辅助（CAD）方法对乳腺病变良恶性鉴别的精度,本文基于多模态特征融合,提出一种联合非对称卷积和超轻子空间注意模块的卷积神经网络AC_Ulsam_CNN.首先,采用迁移学习方法预训练模型,筛选出对乳腺病变良恶性鉴别最为有效的DCE-MRI扫描时序.而后,基于最优扫描时序图像,搭建基于AC_Ulsam_CNN网络的模型,以增强分类模型的特征表达能力和鲁棒性.最后,将影像特征与乳腺影像数据报告和数据系统（BI-RADS）分级、表观扩散系数（ADC）和时间-信号强度曲线（TIC）类型等多模态信息进行特征融合,以进一步提高模型对病灶的预测性能.采用五折交叉验证方法进行模型验证,本文方法获得了0.826的准确率（ACC）和0.877的受试者工作曲线下面积（AUC）.这表明该算法在小样本量数据下可较好区分乳腺病变的良恶性,而基于多模态数据的融合模型也进一步丰富了特征信息,从而提高病灶的检出精度,为乳腺病灶良恶性的自动鉴别诊断提供了新方法.     

残差网络分层融合的高光谱地物分类

高光谱图像具有较高的空间分辨率,蕴含着丰富的空间光谱信息,近年来被广泛用于城市地物分类中。在高光谱图像分类过程中,空间光谱特征的提取直接影响着分类精度;传统的高光谱图像特征提取方法只利用了4或8邻域的像素进行简单卷积处理,因而丢失了大量的复杂、有效信息;卷积神经网络（CNN）虽然可以自动提取空间光谱特征,在保留图像空间信息的同时,简化网络模型,但是,随着网络深度增加,网络分类产生退化现象,而且网络间缺乏相关信息的互补性,从而影响分类精度。该工作引入CNN自动提取空间光谱特征,并且针对CNN深度增加所导致的退化问题,设计了面向地物分类的高光谱特征融合残差网络。首先,为了降低高光谱图像的光谱冗余度,利用PCA提取主要光谱波段;然后,为了逐级提取光谱图像的空间光谱特征,定义了卷积核为16,32,64的低、中、高3层残差网络模块,并利用64个1×1的卷积核对3层特征输出进行卷积,完成维度匹配与特征图融合;接着,对融合后的特征图进行全局平均池化(GAP)生成用于分类的特征向量;最后,引入具有可调节机制的Large-Margin Softmax损失函数,监督模型完成训练过程,实现高光谱图像分类。实验采用Indian Pines,University of Pavia和Salinas地区的高光谱图像来验证方法有效性,设置批次训练的样本集为100,网络训练的初始学习率为0.1,当损失函数稳定后学习率降低为0.001,动量为0.9,权重延迟为0.000 1,最大训练迭代次数为2×104,当3个数据集的样本块像素分别设置为25×25,23×23,27×27,网络深度分别为28,32和28时,3个数据集的分类准确率最高,其平均总体准确率（OA）为98.75%、平均准确率（AA）的评价值为98.1%,平均Kappa系数为0.98。实验结果表明,基于残差网络的分类方法能够自动学习更丰富的空间光谱特征,残差网络层数的增加和不同网络层融合可以提高高光谱分类精度;Large-Margin Softmax实现了类内紧凑和类间分离,可以进一步提高高光谱图像分类精度。     

矢量泰勒级数特征补偿的说话人识别

将矢量泰勒级数(Vector Taylor Series,VTS)特征补偿算法应用于说话人识别,给出了卷积噪声方差的近似闭式解,构建了联合快速估计卷积噪声和加性噪声均值和方差的框架。该算法可在无需失配环境先验信息的前提下,直接从失配语音中估计出卷积噪声和加性噪声的均值和方差,实现对环境失配的补偿。实验结果表明,在信道变化较大的无线信道下,卷积噪声方差的补偿最高可降低误识率3.24%.提升了系统的识别性能。在存在加性噪声的无线信道下,与基于线性失真模型的特征映射算法和倒谱均值减算法相比,本文算法可分别最大降低49.65%和68.06%的误识率,适合于信道变化较大的失配环境补偿。      

基于深度学习的铸件X射线图像分割研究

针对当前图像分割算法在实现工业铸件内部缺陷分割上精度低且算法不够轻量化的问题,提出一种基于改进DeepLabv3+的工业铸件内部缺陷检测算法Effi-DeepLab。该方法采用EfficientNet中的MBConv来代替原有的Xception模块进行特征提取,使特征提取网络更加高效与轻量化;针对工业铸件内部缺陷尺寸小的问题,重新设计空洞空间金字塔池化（ASPP）层中空洞卷积的扩张率,使得卷积块对小目标具有更高的鲁棒性;在解码端充分利用特征提取阶段的低阶语义信息进行多尺度特征融合,以提高小目标缺陷分割的精度。实验结果表明,在本文使用的汽车轮毂内部缺陷图像数据集中,Effi-DeepLab模型对缺陷的分割准确率和平均交并比（mIoU）分别为93.58%和89.39%,相比DeepLabv3+分别提升了2.65%和2.24%,具有更好的分割效果;此外,还通过实验验证了本文提出算法具有良好的泛化性。     

MFF-Net: Deepfake Detection Network Based on Multi-Feature Fusion

Significant progress has been made in generating counterfeit images and videos. Forged videos generated by deepfaking have been widely spread and have caused severe societal impacts, which stir up public concern about automatic deepfake detection technology. Recently, many deepfake detection methods based on forged features have been proposed. Among the popular forged features, textural features are widely used. However, most of the current texture-based detection methods extract textures directly from RGB images, ignoring the mature spectral analysis methods. Therefore, this research proposes a deepfake detection network fusing RGB features and textural information extracted by neural networks and signal processing methods, namely, MFF-Net. Specifically, it consists of four key components: (1) a feature extraction module to further extract textural and frequency information using the Gabor convolution and residual attention blocks; (2) a texture enhancement module to zoom into the subtle textural features in shallow layers; (3) an attention module to force the classifier to focus on the forged part; (4) two instances of feature fusion to firstly fuse textural features from the shallow RGB branch and feature extraction module and then to fuse the textural features and semantic information. Moreover, we further introduce a new diversity loss to force the feature extraction module to learn features of different scales and directions. The experimental results show that MFF-Net has excellent generalization and has achieved state-of-the-art performance on various deepfake datasets.     

Automatic modulation recognition based on a multiscale network with statistical features

Automatic modulation recognition (AMR) can be used in dynamic spectrum access (DSA) techniques to reduce the pressure on spectrum resources. In this paper, we propose a multiscale convolution-based network model called MSNet-SF, which combines traditional statistical features and deep learning (DL) to balance recognition accuracy and complexity. In the model, the feature information is extracted by two multiscale modules, which consist of unit convolution and three different sizes of convolution kernels arranged in parallel. Additionally, the sparse connectivity of unit convolution enables the network to be more lightweight. Five statistical features (four higher-order cumulants (HOCs) and one zero-centered normalized instantaneous magnitude tightness) are also input into the model and are fully fused with the main feature map by multiplication to achieve complementarity of long-term and short-term features. This approach yields a large performance gain at a small cost and greatly reduces the confusion between QAM16 and QAM64. Simulation results in the RML2018.10A dataset show that the average recognition accuracy of the model improved by 4% after adding the statistical features and achieved an accuracy of more than 97% from 12 dB.     

优化FCOS网络复杂果园环境下绿色苹果检测模型

目标果实的精准识别是实现果园测产和机器自动采摘的基本保障.然而受复杂的非结构化果园环境、绿色苹果与枝叶背景颜色接近等因素的影响,制约着可见光谱范围下目标果实的检测精度,给机器视觉识别带来极大挑战.针对复杂果园环境下的不同光照环境和果实姿态,提出一种优化的一阶全卷积(FCOS)神经网络绿色苹果识别模型.首先,新模型在FC...     

基于多尺度空洞卷积的对抗去雾网络

雾天拍摄的图像存在颜色失真、图像细节模糊的问题,对成像设备采集到的图像质量造成了负面印象。针对雾天搜集图像存在的降质问题,提出了一种基于多尺度空洞卷积的对抗去雾网络。去雾网络的生成器由不同空洞率的卷积模块组成,结合多尺度的策略增加感受野并增强去雾效果;判别器采用多个卷积模块构成,用于区分生成的去雾图像与真实无雾图像;通过计算去雾图像和真实无雾图像之间的感知距离,优化图像的纹理结构并减少噪声信号。实验结果显示,提出算法在公开数据集上获得的峰值信噪比值为22.410 dB,结构相似性值为0.844,色差值为10.545。定量和定性评估表明,采用空洞卷积和感知损失技术设计的去雾网络能够有效地恢复图像的颜色信息和纹理结构。