Hurricane Ida ferociously affected many south-eastern and eastern parts of the United States,making it one of the strongest hurricanes in recent years.Advanced forecast and warning tool has been used to track the path...Hurricane Ida ferociously affected many south-eastern and eastern parts of the United States,making it one of the strongest hurricanes in recent years.Advanced forecast and warning tool has been used to track the path of the ex-Hurricane,Ida,as it left New Orleans on its way towards the northeast,accurately predicting significant supercell development above New York City on September 01,2021.This advanced method accurately detected the area with the highest possible level of convective instability with 24-h lead time and even Level 5,devised in the categorical outlooks legend of the system.Therefore,an extreme level implied a very high probability of the local-scale hazard occurring above the NYC.Cloud model output fields(updrafts and downdrafts,wind shear,near-surface convergence,the vertical component of relative vorticity)show the rapid development of a strong supercell storm with rotating updrafts and a mesocyclone.The characteristic hook-shaped echo signature visible in the reflectivity patterns indicates a signal for a highly precipitable(HP)supercell with the possibility of tornado initiation.Open boundary conditions represent a good basis for simulating a tornado that evolved from a supercell storm,initialized with initial data obtained from a real-time simulation in the period when the bow echo and tornado-like signature occurred.Тhe modeled results agree well with the observations.展开更多
Due to the heterogeneity of the structure on a scale-free network, making the betweennesses of all nodes become homogeneous by reassigning the weights of nodes or edges is very difficult. In order to take advantage of...Due to the heterogeneity of the structure on a scale-free network, making the betweennesses of all nodes become homogeneous by reassigning the weights of nodes or edges is very difficult. In order to take advantage of the important effect of high degree nodes on the shortest path communication and preferentially deliver packets by them to increase the probability to destination, an adaptive local routing strategy on a scale-free network is proposed, in which the node adjusts the forwarding probability with the dynamical traffic load (packet queue length) and the degree distribution of neighbouring nodes. The critical queue length of a node is set to be proportional to its degree, and the node with high degree has a larger critical queue length to store and forward more packets. When the queue length of a high degree node is shorter than its critical queue length, it has a higher probability to forward packets. After higher degree nodes are saturated (whose queue lengths are longer than their critical queue lengths), more packets will be delivered by the lower degree nodes around them. The adaptive local routing strategy increases the probability of a packet finding its destination quickly, and improves the transmission capacity on the scale-free network by reducing routing hops. The simulation results show that the transmission capacity of the adaptive local routing strategy is larger than that of three previous local routing strategies.展开更多
The spatial scale(?shing grid) of ?sheries research af fects the observed spatial patterns of?sheries resources such as catch-per-unit-ef fort(CPUE) and ?shing ef fort. We examined the scale impact of high value(HH) c...The spatial scale(?shing grid) of ?sheries research af fects the observed spatial patterns of?sheries resources such as catch-per-unit-ef fort(CPUE) and ?shing ef fort. We examined the scale impact of high value(HH) clusters of the annual ?shing ef fort for Dosidicus gigas of fshore Peru from 2009 to 2012.For a multi-scale analysis, the original commercial ?shery data were tessellated to twelve spatial scales from 6′ to 72′ with an interval of 6′. Under these spatial scales, D. gigas clusters were identi?ed using the Anselin Local Moran's I. Statistics including the number of points, mean CPUE, standard deviation(SD),skewness, kurtosis, area and centroid were calculated for these HH clusters. We found that the z-score of global Moran's I and the number of points for HH clusters follow a power law scaling relationship from2009 to 2012. The mean ef fort and its SD also follow a power law scaling relationship from 2009 to 2012.The skewness follows a linear scaling relationship in 2010 and 2011 but ?uctuates with spatial scale in2009 and 2012; kurtosis follows a logarithmic scale relationship in 2009, 2011 and 2012 but a linear scale relationship in 2010. Cluster area follows a power law scaling relationship in 2010 and 2012, a linear scaling relationship in 2009, and a quadratic scaling relationship in 2011. Based on the peaks of Moran's I indices and the multi-scale analysis, we conclude that the optimum scales are 12′ in 2009 ? 2011 and 6′ in 2012, while the coarsest allowable scales are 48′ in 2009, 2010 and 2012, and 60′ in 2011. Our research provides the best spatial scales for conducting spatial analysis of this pelagic species, and provides a better understanding of scaling behavior for the ?shing ef fort of D. gigas in the of fshore Peruvian waters.展开更多
It is a challenging topic to develop an efficient algorithm for large scale classification problems in many applications of machine learning. In this paper, a hierarchical clustering and fixed-layer local learning (HC...It is a challenging topic to develop an efficient algorithm for large scale classification problems in many applications of machine learning. In this paper, a hierarchical clustering and fixed-layer local learning (HCFLL) based support vector machine(SVM) algorithm is proposed to deal with this problem. Firstly, HCFLL hierarchically clusters a given dataset into a modified clustering feature tree based on the ideas of unsupervised clustering and supervised clustering. Then it locally trains SVM on each labeled subtree at a fixed-layer of the tree. The experimental results show that compared with the existing popular algorithms such as core vector machine and decision-tree support vector machine, HCFLL can significantly improve the training and testing speeds with comparable testing accuracy.展开更多
Hindukush is an active subduction zone where at least one earthquake occurs on daily basis.For seismic hazard studies,it is important to develop a local magnitude scale using the data of local seismic network.We have ...Hindukush is an active subduction zone where at least one earthquake occurs on daily basis.For seismic hazard studies,it is important to develop a local magnitude scale using the data of local seismic network.We have computed local magnitude scale for Hindukush earthquakes using data from local network belonging to Center for Earthquake Studies(CES)for a period of three years,i.e.2015–2017.A total of 26,365 seismic records pertaining to 2,683 earthquakes with magnitude 2.0 and greater,was used with hypocentral distance less than 600 km.Magnitude scale developed by using this data comes to be M_(L)=logA+0.929logr+0.00298r-1.84.The magnitude determined through formulated relation was compared with that of standard relation for Southern California and relation developed by the same authors for local network for Northern Punjab.It was observed that Hindukush region has high attenuation as compared to that of Southern California and Northern Punjab which implies that Hindukush is tectonically more disturbed as compared to the said regions,hence,seismically more active as well.We have calculated station correction factors for our network.Station correction factors do not show any pattern which probably owes to the geological and tectonic complexity of this structure.Standard deviation and variance of magnitude residuals for CES network determined using Hutton and Boore scale and scale developed in this study were compared,it showed that a variance reduction of 44.1%was achieved.Average of magnitude residuals for different distance ranges was almost zero which showed that our magnitude scale was stable for all distances up to 600 km.Newly developed magnitude scale will help in homogenization of earthquake catalog.It has been observed that b-value of CES catalog decreases when magnitude is calculated by using newly developed magnitude scale.展开更多
Developing ecological scale breeding of local chickens with natural conditions, such as forest lands, grass lands, orchards and mulberry fields, not only can improve the quality of poultry products and the production ...Developing ecological scale breeding of local chickens with natural conditions, such as forest lands, grass lands, orchards and mulberry fields, not only can improve the quality of poultry products and the production benefits of chicken breeding, but also can develop and use local chicken germplasm effectively and promote their breed protection and industrialization. From integration and application of breed selection, breeding management, nutrient regulation, grass planting and grazing, a new ecological scale breeding technique of local chickens is proposed.展开更多
Local diversity AdaBoost support vector machine(LDAB-SVM) is proposed for large scale dataset classification problems.The training dataset is split into several blocks firstly, and some models based on these dataset...Local diversity AdaBoost support vector machine(LDAB-SVM) is proposed for large scale dataset classification problems.The training dataset is split into several blocks firstly, and some models based on these dataset blocks are built.In order to obtain a better performance, AdaBoost is used in each model building.In the boosting iteration step, the component learners which have higher diversity and accuracy are collected via the kernel parameters adjusting.Then the local models via voting method are integrated.The experimental study shows that LDAB-SVM can deal with large scale dataset efficiently without reducing the performance of the classifier.展开更多
The traditional description of atomic-scale friction, as investigated in Friction force microscopy, in terms of mechanical stick-slip instabilities appears so successful that it obscures the actual mechanisms of frict...The traditional description of atomic-scale friction, as investigated in Friction force microscopy, in terms of mechanical stick-slip instabilities appears so successful that it obscures the actual mechanisms of frictional energy dissipation. More sophisticated theoretical approach, which takes into account damping explicitly, reveals the existence of some hidden, unexplained problems, like the universal nearly-critical damping and unexpectedly high value of the dissipation rate. In this paper, we combine analysis in the framework of nonequilibrium statistical mechanics with simple atomistic modeling to show that the hidden problems of atomic scale friction find their origin in the nontrivial character of energy dissipation that is non-local and dominated by memory effects, which have not been addressed before in the context of dry, atomic-scale friction.展开更多
Background: The prevalence of carpal tunnel syndrome (CTS) and of anxiety and depression in primary care practice are high. Different studies had shown an increased prevalence of anxiety and depression in CTS patients...Background: The prevalence of carpal tunnel syndrome (CTS) and of anxiety and depression in primary care practice are high. Different studies had shown an increased prevalence of anxiety and depression in CTS patients. Nevertheless, few papers had been published studying the anxiety and depression scales in the treatment of CTS, either with corticosteroid injections (I) or with surgical decompression (S). Objective: To assess whether clinical improvement observed after the treatment of CTS either with I or with S correlates with an improvement in the punctuations of the Hospital Anxiety and Depression scales (HADS), at 3, 6 and 12-month follow-up. Methods: Randomized and open-label clinical trial, comparing I and S. Patients with symptoms suggestive of CTS (nocturnal paraesthesias) of at least 3 months duration and neurophysiological confirmation were included. Patients with clinically apparent motor impairment were excluded. The subjective evaluation of symptoms was carried out using the visual-analogue scale of pain (VAS-p). Clinical reviews were performed 3, 6 and 12 months after treatment. Each patient completed the HADS questionnaire and a VAS-p at 0, 3, 6, and 12 months. Statistical significance was established using the Student’s t test and the Mann-Whitney U test when necessary. A linear regression analysis was used to know the effect of the treatment adjusted for the initial score of both scales. Results: 65 patients were included (30 in group I and 35 in group S). There was no statistical difference between both groups in terms of age, gender distribution, disease duration, VAS-p, neurophysiological testing severity of CTS or the 8 subscales of HADS. Both groups improved significantly in relation to the baseline VAS-p values, in the reviews at 3, 6 and 12 months, with no significant differences between I and S. At 6 months, the reduction in the anxiety scale was around 3 points for both treatments (S = 3.6 and I = 3.2), without reaching significant differences. At 12 months, it was somewhat higher for those treated with I, but always around 3 points and without significant differences. The Depression scale score was slightly reduced at 6 months, and in a similar way for both groups (I = 1 and S = 1.19;p = 0.8). After 12 months, group I doubled the previous reduction, with group S experiencing a very slight change (I = 1.96 and S = 1.03;p = 0.3). When analysing the effect of group S on group I, the result was a reduction of 0.25 points for Anxiety (p = 0.7) and of 0.02 points for Depression (p = 0.9). Conclusions: Treatment of CTS with I or S results in a similar and discrete improvement in Anxiety scores on the HADS scale at 6 and 12 months. For both types of treatment, the Depression scores barely changed at 6 months, being somewhat higher in group I after 12-month follow-up. The independent effect of the S on both scales is small and not significant.展开更多
U-Net在图像分割领域取得了巨大成功,然而卷积和下采样操作导致部分位置信息丢失,全局和长距离的语义交互信息难以被学习,并且缺乏整合全局和局部信息的能力。为了提取丰富的局部细节和全局上下文信息,提出了一个基于卷积胶囊编码器和...U-Net在图像分割领域取得了巨大成功,然而卷积和下采样操作导致部分位置信息丢失,全局和长距离的语义交互信息难以被学习,并且缺乏整合全局和局部信息的能力。为了提取丰富的局部细节和全局上下文信息,提出了一个基于卷积胶囊编码器和局部共现的医学图像分割网络MLFCNet(network based on convolution capsule encoder and multi-scale local feature co-occurrence)。在U-Net基础上引入胶囊网络模块,学习目标位置信息、局部与全局的关系。同时利用提出的注意力机制保留网络池化层丢弃的信息,并且设计了新的多尺度特征融合方法,从而捕捉全局信息并抑制背景噪声。此外,提出了一种新的多尺度局部特征共现算法,局部特征之间的关系能够被更好地学习。在两个公共数据集上与九种方法进行了比较,相比于性能第二的模型,该方法的mIoU在肝脏医学图像中提升了4.7%,Dice系数提升了1.7%。在肝脏医学图像和人像数据集上的实验结果表明,在相同的实验条件下,提出的网络优于U-Net和其他主流的图像分割网络。展开更多
文摘Hurricane Ida ferociously affected many south-eastern and eastern parts of the United States,making it one of the strongest hurricanes in recent years.Advanced forecast and warning tool has been used to track the path of the ex-Hurricane,Ida,as it left New Orleans on its way towards the northeast,accurately predicting significant supercell development above New York City on September 01,2021.This advanced method accurately detected the area with the highest possible level of convective instability with 24-h lead time and even Level 5,devised in the categorical outlooks legend of the system.Therefore,an extreme level implied a very high probability of the local-scale hazard occurring above the NYC.Cloud model output fields(updrafts and downdrafts,wind shear,near-surface convergence,the vertical component of relative vorticity)show the rapid development of a strong supercell storm with rotating updrafts and a mesocyclone.The characteristic hook-shaped echo signature visible in the reflectivity patterns indicates a signal for a highly precipitable(HP)supercell with the possibility of tornado initiation.Open boundary conditions represent a good basis for simulating a tornado that evolved from a supercell storm,initialized with initial data obtained from a real-time simulation in the period when the bow echo and tornado-like signature occurred.Тhe modeled results agree well with the observations.
基金Project supported in part by the National Natural Science Foundation of China (Grant Nos. 60872011 and 60502017)the State Key Development Program for Basic Research of China (Grant Nos. 2009CB320504 and 2010CB731800)Program for New Century Excellent Talents in University
文摘Due to the heterogeneity of the structure on a scale-free network, making the betweennesses of all nodes become homogeneous by reassigning the weights of nodes or edges is very difficult. In order to take advantage of the important effect of high degree nodes on the shortest path communication and preferentially deliver packets by them to increase the probability to destination, an adaptive local routing strategy on a scale-free network is proposed, in which the node adjusts the forwarding probability with the dynamical traffic load (packet queue length) and the degree distribution of neighbouring nodes. The critical queue length of a node is set to be proportional to its degree, and the node with high degree has a larger critical queue length to store and forward more packets. When the queue length of a high degree node is shorter than its critical queue length, it has a higher probability to forward packets. After higher degree nodes are saturated (whose queue lengths are longer than their critical queue lengths), more packets will be delivered by the lower degree nodes around them. The adaptive local routing strategy increases the probability of a packet finding its destination quickly, and improves the transmission capacity on the scale-free network by reducing routing hops. The simulation results show that the transmission capacity of the adaptive local routing strategy is larger than that of three previous local routing strategies.
基金Supported by the National Natural Science Foundation of China(No.41406146)the Laboratory for Marine Fisheries Science and Food Production Processes at Qingdao National Laboratory for Marine Science and Technology of China(No.2017-1A02)the Shanghai Universities First-class Disciplines Project-Fisheries(A)
文摘The spatial scale(?shing grid) of ?sheries research af fects the observed spatial patterns of?sheries resources such as catch-per-unit-ef fort(CPUE) and ?shing ef fort. We examined the scale impact of high value(HH) clusters of the annual ?shing ef fort for Dosidicus gigas of fshore Peru from 2009 to 2012.For a multi-scale analysis, the original commercial ?shery data were tessellated to twelve spatial scales from 6′ to 72′ with an interval of 6′. Under these spatial scales, D. gigas clusters were identi?ed using the Anselin Local Moran's I. Statistics including the number of points, mean CPUE, standard deviation(SD),skewness, kurtosis, area and centroid were calculated for these HH clusters. We found that the z-score of global Moran's I and the number of points for HH clusters follow a power law scaling relationship from2009 to 2012. The mean ef fort and its SD also follow a power law scaling relationship from 2009 to 2012.The skewness follows a linear scaling relationship in 2010 and 2011 but ?uctuates with spatial scale in2009 and 2012; kurtosis follows a logarithmic scale relationship in 2009, 2011 and 2012 but a linear scale relationship in 2010. Cluster area follows a power law scaling relationship in 2010 and 2012, a linear scaling relationship in 2009, and a quadratic scaling relationship in 2011. Based on the peaks of Moran's I indices and the multi-scale analysis, we conclude that the optimum scales are 12′ in 2009 ? 2011 and 6′ in 2012, while the coarsest allowable scales are 48′ in 2009, 2010 and 2012, and 60′ in 2011. Our research provides the best spatial scales for conducting spatial analysis of this pelagic species, and provides a better understanding of scaling behavior for the ?shing ef fort of D. gigas in the of fshore Peruvian waters.
基金National Natural Science Foundation of China ( No. 61070033 )Fundamental Research Funds for the Central Universities,China( No. 2012ZM0061)
文摘It is a challenging topic to develop an efficient algorithm for large scale classification problems in many applications of machine learning. In this paper, a hierarchical clustering and fixed-layer local learning (HCFLL) based support vector machine(SVM) algorithm is proposed to deal with this problem. Firstly, HCFLL hierarchically clusters a given dataset into a modified clustering feature tree based on the ideas of unsupervised clustering and supervised clustering. Then it locally trains SVM on each labeled subtree at a fixed-layer of the tree. The experimental results show that compared with the existing popular algorithms such as core vector machine and decision-tree support vector machine, HCFLL can significantly improve the training and testing speeds with comparable testing accuracy.
文摘Hindukush is an active subduction zone where at least one earthquake occurs on daily basis.For seismic hazard studies,it is important to develop a local magnitude scale using the data of local seismic network.We have computed local magnitude scale for Hindukush earthquakes using data from local network belonging to Center for Earthquake Studies(CES)for a period of three years,i.e.2015–2017.A total of 26,365 seismic records pertaining to 2,683 earthquakes with magnitude 2.0 and greater,was used with hypocentral distance less than 600 km.Magnitude scale developed by using this data comes to be M_(L)=logA+0.929logr+0.00298r-1.84.The magnitude determined through formulated relation was compared with that of standard relation for Southern California and relation developed by the same authors for local network for Northern Punjab.It was observed that Hindukush region has high attenuation as compared to that of Southern California and Northern Punjab which implies that Hindukush is tectonically more disturbed as compared to the said regions,hence,seismically more active as well.We have calculated station correction factors for our network.Station correction factors do not show any pattern which probably owes to the geological and tectonic complexity of this structure.Standard deviation and variance of magnitude residuals for CES network determined using Hutton and Boore scale and scale developed in this study were compared,it showed that a variance reduction of 44.1%was achieved.Average of magnitude residuals for different distance ranges was almost zero which showed that our magnitude scale was stable for all distances up to 600 km.Newly developed magnitude scale will help in homogenization of earthquake catalog.It has been observed that b-value of CES catalog decreases when magnitude is calculated by using newly developed magnitude scale.
基金Supported by Three New Agricultural Project in Jiangsu Province(SXGC[2013]234)
文摘Developing ecological scale breeding of local chickens with natural conditions, such as forest lands, grass lands, orchards and mulberry fields, not only can improve the quality of poultry products and the production benefits of chicken breeding, but also can develop and use local chicken germplasm effectively and promote their breed protection and industrialization. From integration and application of breed selection, breeding management, nutrient regulation, grass planting and grazing, a new ecological scale breeding technique of local chickens is proposed.
基金supported by the National Natural Science Foundation of China (60603098)
文摘Local diversity AdaBoost support vector machine(LDAB-SVM) is proposed for large scale dataset classification problems.The training dataset is split into several blocks firstly, and some models based on these dataset blocks are built.In order to obtain a better performance, AdaBoost is used in each model building.In the boosting iteration step, the component learners which have higher diversity and accuracy are collected via the kernel parameters adjusting.Then the local models via voting method are integrated.The experimental study shows that LDAB-SVM can deal with large scale dataset efficiently without reducing the performance of the classifier.
文摘The traditional description of atomic-scale friction, as investigated in Friction force microscopy, in terms of mechanical stick-slip instabilities appears so successful that it obscures the actual mechanisms of frictional energy dissipation. More sophisticated theoretical approach, which takes into account damping explicitly, reveals the existence of some hidden, unexplained problems, like the universal nearly-critical damping and unexpectedly high value of the dissipation rate. In this paper, we combine analysis in the framework of nonequilibrium statistical mechanics with simple atomistic modeling to show that the hidden problems of atomic scale friction find their origin in the nontrivial character of energy dissipation that is non-local and dominated by memory effects, which have not been addressed before in the context of dry, atomic-scale friction.
文摘识别非驾驶行为是提高驾驶安全性的重要手段之一。目前基于骨架序列和图像的融合识别方法具有计算量大和特征融合困难的问题。针对上述问题,本文提出一种基于多尺度骨架图和局部视觉上下文融合的驾驶员行为识别模型(skeleton-image based behavior recognition network,SIBBR-Net)。SIBBR-Net通过基于多尺度图的图卷积网络和基于局部视觉及注意力机制的卷积神经网络,充分提取运动和外观特征,较好地平衡了模型表征能力和计算量间的关系。基于手部运动的特征双向引导学习策略、自适应特征融合模块和静态特征空间上的辅助损失,使运动和外观特征间互相引导更新并实现自适应融合。最终在Drive&Act数据集进行算法测试,SIBBR-Net在动态标签和静态标签条件下的平均正确率分别为61.78%和80.42%,每秒浮点运算次数为25.92G,较最优方法降低了76.96%。
文摘Background: The prevalence of carpal tunnel syndrome (CTS) and of anxiety and depression in primary care practice are high. Different studies had shown an increased prevalence of anxiety and depression in CTS patients. Nevertheless, few papers had been published studying the anxiety and depression scales in the treatment of CTS, either with corticosteroid injections (I) or with surgical decompression (S). Objective: To assess whether clinical improvement observed after the treatment of CTS either with I or with S correlates with an improvement in the punctuations of the Hospital Anxiety and Depression scales (HADS), at 3, 6 and 12-month follow-up. Methods: Randomized and open-label clinical trial, comparing I and S. Patients with symptoms suggestive of CTS (nocturnal paraesthesias) of at least 3 months duration and neurophysiological confirmation were included. Patients with clinically apparent motor impairment were excluded. The subjective evaluation of symptoms was carried out using the visual-analogue scale of pain (VAS-p). Clinical reviews were performed 3, 6 and 12 months after treatment. Each patient completed the HADS questionnaire and a VAS-p at 0, 3, 6, and 12 months. Statistical significance was established using the Student’s t test and the Mann-Whitney U test when necessary. A linear regression analysis was used to know the effect of the treatment adjusted for the initial score of both scales. Results: 65 patients were included (30 in group I and 35 in group S). There was no statistical difference between both groups in terms of age, gender distribution, disease duration, VAS-p, neurophysiological testing severity of CTS or the 8 subscales of HADS. Both groups improved significantly in relation to the baseline VAS-p values, in the reviews at 3, 6 and 12 months, with no significant differences between I and S. At 6 months, the reduction in the anxiety scale was around 3 points for both treatments (S = 3.6 and I = 3.2), without reaching significant differences. At 12 months, it was somewhat higher for those treated with I, but always around 3 points and without significant differences. The Depression scale score was slightly reduced at 6 months, and in a similar way for both groups (I = 1 and S = 1.19;p = 0.8). After 12 months, group I doubled the previous reduction, with group S experiencing a very slight change (I = 1.96 and S = 1.03;p = 0.3). When analysing the effect of group S on group I, the result was a reduction of 0.25 points for Anxiety (p = 0.7) and of 0.02 points for Depression (p = 0.9). Conclusions: Treatment of CTS with I or S results in a similar and discrete improvement in Anxiety scores on the HADS scale at 6 and 12 months. For both types of treatment, the Depression scores barely changed at 6 months, being somewhat higher in group I after 12-month follow-up. The independent effect of the S on both scales is small and not significant.
文摘U-Net在图像分割领域取得了巨大成功,然而卷积和下采样操作导致部分位置信息丢失,全局和长距离的语义交互信息难以被学习,并且缺乏整合全局和局部信息的能力。为了提取丰富的局部细节和全局上下文信息,提出了一个基于卷积胶囊编码器和局部共现的医学图像分割网络MLFCNet(network based on convolution capsule encoder and multi-scale local feature co-occurrence)。在U-Net基础上引入胶囊网络模块,学习目标位置信息、局部与全局的关系。同时利用提出的注意力机制保留网络池化层丢弃的信息,并且设计了新的多尺度特征融合方法,从而捕捉全局信息并抑制背景噪声。此外,提出了一种新的多尺度局部特征共现算法,局部特征之间的关系能够被更好地学习。在两个公共数据集上与九种方法进行了比较,相比于性能第二的模型,该方法的mIoU在肝脏医学图像中提升了4.7%,Dice系数提升了1.7%。在肝脏医学图像和人像数据集上的实验结果表明,在相同的实验条件下,提出的网络优于U-Net和其他主流的图像分割网络。