Convolutional neural networks (CNNs) are widely used in image classification tasks, but their increasing model size and computation make them challenging to implement on embedded systems with constrained hardware reso...Convolutional neural networks (CNNs) are widely used in image classification tasks, but their increasing model size and computation make them challenging to implement on embedded systems with constrained hardware resources. To address this issue, the MobileNetV1 network was developed, which employs depthwise convolution to reduce network complexity. MobileNetV1 employs a stride of 2 in several convolutional layers to decrease the spatial resolution of feature maps, thereby lowering computational costs. However, this stride setting can lead to a loss of spatial information, particularly affecting the detection and representation of smaller objects or finer details in images. To maintain the trade-off between complexity and model performance, a lightweight convolutional neural network with hierarchical multi-scale feature fusion based on the MobileNetV1 network is proposed. The network consists of two main subnetworks. The first subnetwork uses a depthwise dilated separable convolution (DDSC) layer to learn imaging features with fewer parameters, which results in a lightweight and computationally inexpensive network. Furthermore, depthwise dilated convolution in DDSC layer effectively expands the field of view of filters, allowing them to incorporate a larger context. The second subnetwork is a hierarchical multi-scale feature fusion (HMFF) module that uses parallel multi-resolution branches architecture to process the input feature map in order to extract the multi-scale feature information of the input image. Experimental results on the CIFAR-10, Malaria, and KvasirV1 datasets demonstrate that the proposed method is efficient, reducing the network parameters and computational cost by 65.02% and 39.78%, respectively, while maintaining the network performance compared to the MobileNetV1 baseline.展开更多
针对小目标水漂垃圾形态多变、分辨率低且信息有限,导致检测效果不理想的问题,提出一种改进的Faster-RCNN(Faster Regions with Convolutional Neural Network)水漂垃圾检测算法MP-Faster-RCNN(Faster-RCNN with Multi-scale feature an...针对小目标水漂垃圾形态多变、分辨率低且信息有限,导致检测效果不理想的问题,提出一种改进的Faster-RCNN(Faster Regions with Convolutional Neural Network)水漂垃圾检测算法MP-Faster-RCNN(Faster-RCNN with Multi-scale feature and Polarized self-attention)。首先,建立黄河兰州段小目标水漂垃圾数据集,将空洞卷积结合ResNet-50代替原来的VGG-16(Visual Geometry Group 16)作为主干特征提取网络,扩大感受野以提取更多小目标特征;其次,在区域生成网络(RPN)利用多尺度特征,设置3×3和1×1的两层卷积,补偿单一滑动窗口造成的特征丢失;最后,在RPN前加入极化自注意力,进一步利用多尺度和通道特征提取更细粒度的多尺度空间信息和通道间依赖关系,生成具有全局特征的特征图,实现更精确的目标框定位。实验结果表明,MP-Faster-RCNN能有效提高水漂垃圾检测精度,与原始Faster-RCNN相比,平均精度均值(mAP)提高了6.37个百分点,模型大小从521 MB降到了108 MB,且在同一训练批次下收敛更快。展开更多
针对Yolov3-Tiny算法在加油站监控场景检测时由于数据特征提取不充分而导致检测精度低、漏检率高等问题,提出一种基于加油站场景的Misp-YOLO(You Only Look Once)目标检测算法。首先引入Mosaic数据增强算法,使图片包含更多特征信息;其...针对Yolov3-Tiny算法在加油站监控场景检测时由于数据特征提取不充分而导致检测精度低、漏检率高等问题,提出一种基于加油站场景的Misp-YOLO(You Only Look Once)目标检测算法。首先引入Mosaic数据增强算法,使图片包含更多特征信息;其次使用InceptionV2和PSConv(Poly-Scale Convolution)多尺度特征提取方法提升网络多尺度预测能力;最后结合scSE(Concurrent Spatial and Channel ‘Squeeze&Excitation’)注意力机制,重构主干网络输出特征。实验结果证明该算法具有较高检测准确度,并且检测速度满足实际需求。优化后的算法性能得到极大提升,可推广应用于其他目标检测中。展开更多
场地-城市相互作用(site-city interaction,SCI)效应会显著改变场地地震波场分布及建筑反应,基于SCI效应理论计算研究方法的发展现状,发挥谱元(spectral element,SE)法可快速高效求解三维地震波场传播和多自由度(multi-degree of freedo...场地-城市相互作用(site-city interaction,SCI)效应会显著改变场地地震波场分布及建筑反应,基于SCI效应理论计算研究方法的发展现状,发挥谱元(spectral element,SE)法可快速高效求解三维地震波场传播和多自由度(multi-degree of freedom,MDOF)模型计算量小且可同时模拟大量建筑的优势,同时,结合频率波数域(frequency wave number analysis,FK)方法,以等效地震荷载的方式施加地震波场,建立了FK-SE-MDOF耦合方法,实现了SE-MDOF耦合模型中多种波型(P波、SV波和SH波)的斜入射输入,解决了当前三维SCI效应研究方法中未能同时考虑建筑非线性、频谱特性、地震波波型及入射角度影响的问题。首先对方法原理进行了介绍;然后,通过与振动台试验的对比,验证了方法的正确性;进而,采用该方法建立理想场地-城市建筑群相互作用耦合模型,主要探讨了入射角度和地震波波型对SCI效应的影响,得到了一些有益结论。该方法较为真实地反映SCI效应影响的同时,可反映建筑基础轮廓对地震波场的影响,适用于需考虑建筑轮廓信息的社区尺度SCI效应研究,可为城市规划、抗震设计、风险评估以及震后救援等工作提供定量指导。展开更多
文摘Convolutional neural networks (CNNs) are widely used in image classification tasks, but their increasing model size and computation make them challenging to implement on embedded systems with constrained hardware resources. To address this issue, the MobileNetV1 network was developed, which employs depthwise convolution to reduce network complexity. MobileNetV1 employs a stride of 2 in several convolutional layers to decrease the spatial resolution of feature maps, thereby lowering computational costs. However, this stride setting can lead to a loss of spatial information, particularly affecting the detection and representation of smaller objects or finer details in images. To maintain the trade-off between complexity and model performance, a lightweight convolutional neural network with hierarchical multi-scale feature fusion based on the MobileNetV1 network is proposed. The network consists of two main subnetworks. The first subnetwork uses a depthwise dilated separable convolution (DDSC) layer to learn imaging features with fewer parameters, which results in a lightweight and computationally inexpensive network. Furthermore, depthwise dilated convolution in DDSC layer effectively expands the field of view of filters, allowing them to incorporate a larger context. The second subnetwork is a hierarchical multi-scale feature fusion (HMFF) module that uses parallel multi-resolution branches architecture to process the input feature map in order to extract the multi-scale feature information of the input image. Experimental results on the CIFAR-10, Malaria, and KvasirV1 datasets demonstrate that the proposed method is efficient, reducing the network parameters and computational cost by 65.02% and 39.78%, respectively, while maintaining the network performance compared to the MobileNetV1 baseline.
文摘由于低照度图像具有对比度低、细节丢失严重、噪声大等缺点,现有的目标检测算法对低照度图像的检测效果不理想.为此,本文提出一种结合空间感知注意力机制和多尺度特征融合(Spatial-aware Attention Mechanism and Multi-Scale Feature Fusion,SAM-MSFF)的低照度目标检测方法 .该方法首先通过多尺度交互内存金字塔融合多尺度特征,增强低照度图像特征中的有效信息,并设置内存向量存储样本的特征,捕获样本之间的潜在关联性;然后,引入空间感知注意力机制获取特征在空间域的长距离上下文信息和局部信息,从而增强低照度图像中的目标特征,抑制背景信息和噪声的干扰;最后,利用多感受野增强模块扩张特征的感受野,对具有不同感受野的特征进行分组重加权计算,使检测网络根据输入的多尺度信息自适应地调整感受野的大小.在ExDark数据集上进行实验,本文方法的平均精度(mean Average Precision,mAP)达到77.04%,比现有的主流目标检测方法提高2.6%~14.34%.
文摘针对小目标水漂垃圾形态多变、分辨率低且信息有限,导致检测效果不理想的问题,提出一种改进的Faster-RCNN(Faster Regions with Convolutional Neural Network)水漂垃圾检测算法MP-Faster-RCNN(Faster-RCNN with Multi-scale feature and Polarized self-attention)。首先,建立黄河兰州段小目标水漂垃圾数据集,将空洞卷积结合ResNet-50代替原来的VGG-16(Visual Geometry Group 16)作为主干特征提取网络,扩大感受野以提取更多小目标特征;其次,在区域生成网络(RPN)利用多尺度特征,设置3×3和1×1的两层卷积,补偿单一滑动窗口造成的特征丢失;最后,在RPN前加入极化自注意力,进一步利用多尺度和通道特征提取更细粒度的多尺度空间信息和通道间依赖关系,生成具有全局特征的特征图,实现更精确的目标框定位。实验结果表明,MP-Faster-RCNN能有效提高水漂垃圾检测精度,与原始Faster-RCNN相比,平均精度均值(mAP)提高了6.37个百分点,模型大小从521 MB降到了108 MB,且在同一训练批次下收敛更快。
文摘针对Yolov3-Tiny算法在加油站监控场景检测时由于数据特征提取不充分而导致检测精度低、漏检率高等问题,提出一种基于加油站场景的Misp-YOLO(You Only Look Once)目标检测算法。首先引入Mosaic数据增强算法,使图片包含更多特征信息;其次使用InceptionV2和PSConv(Poly-Scale Convolution)多尺度特征提取方法提升网络多尺度预测能力;最后结合scSE(Concurrent Spatial and Channel ‘Squeeze&Excitation’)注意力机制,重构主干网络输出特征。实验结果证明该算法具有较高检测准确度,并且检测速度满足实际需求。优化后的算法性能得到极大提升,可推广应用于其他目标检测中。
文摘场地-城市相互作用(site-city interaction,SCI)效应会显著改变场地地震波场分布及建筑反应,基于SCI效应理论计算研究方法的发展现状,发挥谱元(spectral element,SE)法可快速高效求解三维地震波场传播和多自由度(multi-degree of freedom,MDOF)模型计算量小且可同时模拟大量建筑的优势,同时,结合频率波数域(frequency wave number analysis,FK)方法,以等效地震荷载的方式施加地震波场,建立了FK-SE-MDOF耦合方法,实现了SE-MDOF耦合模型中多种波型(P波、SV波和SH波)的斜入射输入,解决了当前三维SCI效应研究方法中未能同时考虑建筑非线性、频谱特性、地震波波型及入射角度影响的问题。首先对方法原理进行了介绍;然后,通过与振动台试验的对比,验证了方法的正确性;进而,采用该方法建立理想场地-城市建筑群相互作用耦合模型,主要探讨了入射角度和地震波波型对SCI效应的影响,得到了一些有益结论。该方法较为真实地反映SCI效应影响的同时,可反映建筑基础轮廓对地震波场的影响,适用于需考虑建筑轮廓信息的社区尺度SCI效应研究,可为城市规划、抗震设计、风险评估以及震后救援等工作提供定量指导。