The current existing problem of deep learning framework for the detection and segmentation of electrical equipment is dominantly related to low precision.Because of the reliable,safe and easy-to-operate technology pro...The current existing problem of deep learning framework for the detection and segmentation of electrical equipment is dominantly related to low precision.Because of the reliable,safe and easy-to-operate technology provided by deep learning-based video surveillance for unmanned inspection of electrical equipment,this paper uses the bottleneck attention module(BAM)attention mechanism to improve the Solov2 model and proposes a new electrical equipment segmentation mode.Firstly,the BAM attention mechanism is integrated into the feature extraction network to adaptively learn the correlation between feature channels,thereby improving the expression ability of the feature map;secondly,the weighted sum of CrossEntropy Loss and Dice loss is designed as the mask loss to improve the segmentation accuracy and robustness of the model;finally,the non-maximal suppression(NMS)algorithm to better handle the overlap problem in instance segmentation.Experimental results show that the proposed method achieves an average segmentation accuracy of mAP of 80.4% on three types of electrical equipment datasets,including transformers,insulators and voltage transformers,which improve the detection accuracy by more than 5.7% compared with the original Solov2 model.The segmentation model proposed can provide a focusing technical means for the intelligent management of power systems.展开更多
AIM: To explore the effects and mechanism of action of antidepressant mirtazapine in functional dyspepsia(FD) patients with weight loss.METHODS: Sixty depressive FD patients with weight loss were randomly divided into...AIM: To explore the effects and mechanism of action of antidepressant mirtazapine in functional dyspepsia(FD) patients with weight loss.METHODS: Sixty depressive FD patients with weight loss were randomly divided into a mirtazapine group(MG), a paroxetine group(PG) or a conventional therapy group(CG) for an 8-wk clinical trial. Adverse effects and treatment response were recorded. The Nepean Dyspepsia Index-symptom(NDSI) checklist and the 17-item Hamilton Rating Scale of Depression(HAMD-17) were used to evaluate dyspepsia and depressive symptoms, respectively. The body composition analyzer was used to measure body weight and fat. Serum hormone levels were measured by ELISA.RESULTS:(1) After 2 wk of treatment, NDSI scores were significantly lower for the MG than for the PG and CG;(2) After 4 or 8 wk of treatment, HAMD-17 scores were significantly lower for the MG and PG than for the CG;(3) After 8 wk of treatment, patients in the MG experienced a weight gain of 3.58 ± 1.57 kg, which was significantly higher than that observed for patients in the PG and CG. Body fat increased by 2.77 ± 0.14kg, the body fat ratio rose by 4%, and the visceral fat area increased by 7.56 ± 2.25 cm2; and(4) For the MG, serum hormone levels of ghrelin, neuropeptide Y(NPY), motilin(MTL) and gastrin(GAS) were significantly upregulated; in contrast, those of leptin, 5-hydroxytryptamine(5-HT) and cholecystokinin(CCK) were significantly downregulated. CONCLUSION: Mirtazapine not only alleviates symptoms associated with dyspepsia and depression linked to FD in patients with weight loss but also significantly increases body weight(mainly the visceral fat in body fat). The likely mechanism of mirtazapine action is regulation of brain-gut or gastrointestinal hormone levels.展开更多
The multiple patterns of internal solitary wave interactions(ISWI)are a complex oceanic phenomenon.Satellite remote sensing techniques indirectly detect these ISWI,but do not provide information on their detailed stru...The multiple patterns of internal solitary wave interactions(ISWI)are a complex oceanic phenomenon.Satellite remote sensing techniques indirectly detect these ISWI,but do not provide information on their detailed structure and dynamics.Recently,the authors considered a three-layer fluid with shear flow and developed a(2+1)Kadomtsev-Petviashvili(KP)model that is capable of describing five types of oceanic ISWI,including O-type,P-type,TO-type,TP-type,and Y-shaped.Deep learning models,particularly physics-informed neural networks(PINN),are widely used in the field of fluids and internal solitary waves.However,the authors find that the amplitude of internal solitary waves is much smaller than the wavelength and the ISWI occur at relatively large spatial scales,and these characteristics lead to an imbalance in the loss function of the PINN model.To solve this problem,the authors introduce two weighted loss function methods,the fixed weighing and the adaptive weighting methods,to improve the PINN model.This successfully simulated the detailed structure and dynamics of ISWI,with simulation results corresponding to the satellite images.In particular,the adaptive weighting method can automatically update the weights of different terms in the loss function and outperforms the fixed weighting method in terms of generalization ability.展开更多
工业数据由于技术故障和人为因素通常导致数据异常,现有基于约束的方法因约束阈值设置的过于宽松或严格会导致修复错误,基于统计的方法因平滑修复机制导致对时间步长较远的异常值修复准确度较低.针对上述问题,提出了基于奖励机制的最小...工业数据由于技术故障和人为因素通常导致数据异常,现有基于约束的方法因约束阈值设置的过于宽松或严格会导致修复错误,基于统计的方法因平滑修复机制导致对时间步长较远的异常值修复准确度较低.针对上述问题,提出了基于奖励机制的最小迭代修复和改进WGAN混合模型的时序数据修复方法.首先,在预处理阶段,保留异常数据,进行信息标注等处理,从而充分挖掘异常值与真实值之间的特征约束.其次,在噪声模块提出了近邻参数裁剪规则,用于修正最小迭代修复公式生成的噪声向量.将其传递至模拟分布模块的生成器中,同时设计了一个动态时间注意力网络层,用于提取时序特征权重并与门控循环单元串联组合捕捉不同步长的特征依赖,并引入递归多步预测原理共同提升模型的表达能力;在判别器中设计了Abnormal and Truth奖励机制和Weighted Mean Square Error损失函数共同反向优化生成器修复数据的细节和质量.最后,在公开数据集和真实数据集上的实验结果表明,该方法的修复准确度与模型稳定性显著优于现有方法.展开更多
为了更全面地对睡眠脑电进行特征提取,提出一种基于多视图与注意力机制的睡眠脑电分期方法。首先针对原始睡眠脑电信号构造时域和时频域两类视图数据;然后设计融合注意力机制的混合神经网络对多视图数据进行表征学习;接着通过双向长短...为了更全面地对睡眠脑电进行特征提取,提出一种基于多视图与注意力机制的睡眠脑电分期方法。首先针对原始睡眠脑电信号构造时域和时频域两类视图数据;然后设计融合注意力机制的混合神经网络对多视图数据进行表征学习;接着通过双向长短时记忆(bi-directional long short-term memory,BiLSTM)网络进一步学习睡眠阶段之间的转换规则;最后使用Softmax函数进行睡眠分期,并利用类别加权损失函数解决睡眠数据类别不均衡的问题。实验使用Sleep-EDF数据库中前20名受试者的单通道脑电信号并采用20折交叉验证对模型进行性能评估,睡眠分期准确率达到83.7%,宏平均F_(1)值达到79.0%,Cohen′s Kappa系数达到0.78。与现有方法相比,算法性能提升明显,证明了所提方法的有效性。展开更多
Nuclearmagnetic resonance imaging of breasts often presents complex backgrounds.Breast tumors exhibit varying sizes,uneven intensity,and indistinct boundaries.These characteristics can lead to challenges such as low a...Nuclearmagnetic resonance imaging of breasts often presents complex backgrounds.Breast tumors exhibit varying sizes,uneven intensity,and indistinct boundaries.These characteristics can lead to challenges such as low accuracy and incorrect segmentation during tumor segmentation.Thus,we propose a two-stage breast tumor segmentation method leveraging multi-scale features and boundary attention mechanisms.Initially,the breast region of interest is extracted to isolate the breast area from surrounding tissues and organs.Subsequently,we devise a fusion network incorporatingmulti-scale features and boundary attentionmechanisms for breast tumor segmentation.We incorporate multi-scale parallel dilated convolution modules into the network,enhancing its capability to segment tumors of various sizes through multi-scale convolution and novel fusion techniques.Additionally,attention and boundary detection modules are included to augment the network’s capacity to locate tumors by capturing nonlocal dependencies in both spatial and channel domains.Furthermore,a hybrid loss function with boundary weight is employed to address sample class imbalance issues and enhance the network’s boundary maintenance capability through additional loss.Themethod was evaluated using breast data from 207 patients at RuijinHospital,resulting in a 6.64%increase in Dice similarity coefficient compared to the benchmarkU-Net.Experimental results demonstrate the superiority of the method over other segmentation techniques,with fewer model parameters.展开更多
针对现有非机动车头盔佩戴检测算法在车流密集场景中存在漏检,对佩戴其他帽子存在误检的问题,提出一种改进YOLOv5s(you only look once version5)的头盔佩戴检测算法YOLOv5s-BC。首先,采用软池化替换特征金字塔池化结构中的最大池化层,...针对现有非机动车头盔佩戴检测算法在车流密集场景中存在漏检,对佩戴其他帽子存在误检的问题,提出一种改进YOLOv5s(you only look once version5)的头盔佩戴检测算法YOLOv5s-BC。首先,采用软池化替换特征金字塔池化结构中的最大池化层,以放大更大强度的特征激活;其次,将坐标注意力机制和加权双向特征金字塔网络结合,搭建一种高效的双向跨尺度连接的加权特征聚合网络,以增强不同层级之间的信息传播;最后,用EIoU损失函数优化边框回归,精确目标定位。实验结果表明:在自制头盔数据集上,改进后的算法的平均精度(mAP)可达98.4%,比原算法提高了6.3%,推理速度达到58.69帧/s,整体性能优于其他主流算法,可满足交通道路环境下头盔佩戴检测的准确率和实时性要求。展开更多
基金Jilin Science and Technology Development Plan Project(No.20200403075SF)Doctoral Research Start-Up Fund of Northeast Electric Power University(No.BSJXM-2018202).
文摘The current existing problem of deep learning framework for the detection and segmentation of electrical equipment is dominantly related to low precision.Because of the reliable,safe and easy-to-operate technology provided by deep learning-based video surveillance for unmanned inspection of electrical equipment,this paper uses the bottleneck attention module(BAM)attention mechanism to improve the Solov2 model and proposes a new electrical equipment segmentation mode.Firstly,the BAM attention mechanism is integrated into the feature extraction network to adaptively learn the correlation between feature channels,thereby improving the expression ability of the feature map;secondly,the weighted sum of CrossEntropy Loss and Dice loss is designed as the mask loss to improve the segmentation accuracy and robustness of the model;finally,the non-maximal suppression(NMS)algorithm to better handle the overlap problem in instance segmentation.Experimental results show that the proposed method achieves an average segmentation accuracy of mAP of 80.4% on three types of electrical equipment datasets,including transformers,insulators and voltage transformers,which improve the detection accuracy by more than 5.7% compared with the original Solov2 model.The segmentation model proposed can provide a focusing technical means for the intelligent management of power systems.
文摘AIM: To explore the effects and mechanism of action of antidepressant mirtazapine in functional dyspepsia(FD) patients with weight loss.METHODS: Sixty depressive FD patients with weight loss were randomly divided into a mirtazapine group(MG), a paroxetine group(PG) or a conventional therapy group(CG) for an 8-wk clinical trial. Adverse effects and treatment response were recorded. The Nepean Dyspepsia Index-symptom(NDSI) checklist and the 17-item Hamilton Rating Scale of Depression(HAMD-17) were used to evaluate dyspepsia and depressive symptoms, respectively. The body composition analyzer was used to measure body weight and fat. Serum hormone levels were measured by ELISA.RESULTS:(1) After 2 wk of treatment, NDSI scores were significantly lower for the MG than for the PG and CG;(2) After 4 or 8 wk of treatment, HAMD-17 scores were significantly lower for the MG and PG than for the CG;(3) After 8 wk of treatment, patients in the MG experienced a weight gain of 3.58 ± 1.57 kg, which was significantly higher than that observed for patients in the PG and CG. Body fat increased by 2.77 ± 0.14kg, the body fat ratio rose by 4%, and the visceral fat area increased by 7.56 ± 2.25 cm2; and(4) For the MG, serum hormone levels of ghrelin, neuropeptide Y(NPY), motilin(MTL) and gastrin(GAS) were significantly upregulated; in contrast, those of leptin, 5-hydroxytryptamine(5-HT) and cholecystokinin(CCK) were significantly downregulated. CONCLUSION: Mirtazapine not only alleviates symptoms associated with dyspepsia and depression linked to FD in patients with weight loss but also significantly increases body weight(mainly the visceral fat in body fat). The likely mechanism of mirtazapine action is regulation of brain-gut or gastrointestinal hormone levels.
基金supported by the National Natural Science Foundation of China under Grant Nos.12275085,12235007,and 12175069Science and Technology Commission of Shanghai Municipality under Grant Nos.21JC1402500 and 22DZ2229014.
文摘The multiple patterns of internal solitary wave interactions(ISWI)are a complex oceanic phenomenon.Satellite remote sensing techniques indirectly detect these ISWI,but do not provide information on their detailed structure and dynamics.Recently,the authors considered a three-layer fluid with shear flow and developed a(2+1)Kadomtsev-Petviashvili(KP)model that is capable of describing five types of oceanic ISWI,including O-type,P-type,TO-type,TP-type,and Y-shaped.Deep learning models,particularly physics-informed neural networks(PINN),are widely used in the field of fluids and internal solitary waves.However,the authors find that the amplitude of internal solitary waves is much smaller than the wavelength and the ISWI occur at relatively large spatial scales,and these characteristics lead to an imbalance in the loss function of the PINN model.To solve this problem,the authors introduce two weighted loss function methods,the fixed weighing and the adaptive weighting methods,to improve the PINN model.This successfully simulated the detailed structure and dynamics of ISWI,with simulation results corresponding to the satellite images.In particular,the adaptive weighting method can automatically update the weights of different terms in the loss function and outperforms the fixed weighting method in terms of generalization ability.
文摘工业数据由于技术故障和人为因素通常导致数据异常,现有基于约束的方法因约束阈值设置的过于宽松或严格会导致修复错误,基于统计的方法因平滑修复机制导致对时间步长较远的异常值修复准确度较低.针对上述问题,提出了基于奖励机制的最小迭代修复和改进WGAN混合模型的时序数据修复方法.首先,在预处理阶段,保留异常数据,进行信息标注等处理,从而充分挖掘异常值与真实值之间的特征约束.其次,在噪声模块提出了近邻参数裁剪规则,用于修正最小迭代修复公式生成的噪声向量.将其传递至模拟分布模块的生成器中,同时设计了一个动态时间注意力网络层,用于提取时序特征权重并与门控循环单元串联组合捕捉不同步长的特征依赖,并引入递归多步预测原理共同提升模型的表达能力;在判别器中设计了Abnormal and Truth奖励机制和Weighted Mean Square Error损失函数共同反向优化生成器修复数据的细节和质量.最后,在公开数据集和真实数据集上的实验结果表明,该方法的修复准确度与模型稳定性显著优于现有方法.
文摘为了更全面地对睡眠脑电进行特征提取,提出一种基于多视图与注意力机制的睡眠脑电分期方法。首先针对原始睡眠脑电信号构造时域和时频域两类视图数据;然后设计融合注意力机制的混合神经网络对多视图数据进行表征学习;接着通过双向长短时记忆(bi-directional long short-term memory,BiLSTM)网络进一步学习睡眠阶段之间的转换规则;最后使用Softmax函数进行睡眠分期,并利用类别加权损失函数解决睡眠数据类别不均衡的问题。实验使用Sleep-EDF数据库中前20名受试者的单通道脑电信号并采用20折交叉验证对模型进行性能评估,睡眠分期准确率达到83.7%,宏平均F_(1)值达到79.0%,Cohen′s Kappa系数达到0.78。与现有方法相比,算法性能提升明显,证明了所提方法的有效性。
基金funded by the National Natural Foundation of China under Grant No.61172167the Science Fund Project of Heilongjiang Province(LH2020F035).
文摘Nuclearmagnetic resonance imaging of breasts often presents complex backgrounds.Breast tumors exhibit varying sizes,uneven intensity,and indistinct boundaries.These characteristics can lead to challenges such as low accuracy and incorrect segmentation during tumor segmentation.Thus,we propose a two-stage breast tumor segmentation method leveraging multi-scale features and boundary attention mechanisms.Initially,the breast region of interest is extracted to isolate the breast area from surrounding tissues and organs.Subsequently,we devise a fusion network incorporatingmulti-scale features and boundary attentionmechanisms for breast tumor segmentation.We incorporate multi-scale parallel dilated convolution modules into the network,enhancing its capability to segment tumors of various sizes through multi-scale convolution and novel fusion techniques.Additionally,attention and boundary detection modules are included to augment the network’s capacity to locate tumors by capturing nonlocal dependencies in both spatial and channel domains.Furthermore,a hybrid loss function with boundary weight is employed to address sample class imbalance issues and enhance the network’s boundary maintenance capability through additional loss.Themethod was evaluated using breast data from 207 patients at RuijinHospital,resulting in a 6.64%increase in Dice similarity coefficient compared to the benchmarkU-Net.Experimental results demonstrate the superiority of the method over other segmentation techniques,with fewer model parameters.
文摘针对现有非机动车头盔佩戴检测算法在车流密集场景中存在漏检,对佩戴其他帽子存在误检的问题,提出一种改进YOLOv5s(you only look once version5)的头盔佩戴检测算法YOLOv5s-BC。首先,采用软池化替换特征金字塔池化结构中的最大池化层,以放大更大强度的特征激活;其次,将坐标注意力机制和加权双向特征金字塔网络结合,搭建一种高效的双向跨尺度连接的加权特征聚合网络,以增强不同层级之间的信息传播;最后,用EIoU损失函数优化边框回归,精确目标定位。实验结果表明:在自制头盔数据集上,改进后的算法的平均精度(mAP)可达98.4%,比原算法提高了6.3%,推理速度达到58.69帧/s,整体性能优于其他主流算法,可满足交通道路环境下头盔佩戴检测的准确率和实时性要求。
