Bus and any other public transit connectivity issues facilitate an understanding of the importance of transit planning in enhancing existing or new transit services.Improving transit connectivity is one of the most vi...Bus and any other public transit connectivity issues facilitate an understanding of the importance of transit planning in enhancing existing or new transit services.Improving transit connectivity is one of the most vital tasks in transit-operations planning.A poor connection can cause some passengers to stop using the transit service.Service-design criteria always contain postulates to improve routing and scheduling coordination(intra-and inter-agency transfer centers/points and synchronized/timed transfers).Ostensibly the lack of well-defined connectivity measures precludes the weighing and quantifying of the result of any coordination effort.This work provides an initial methodological framework and concepts for(1)quantifying transit connectivity measures and(2)directions and tools for detecting weak segments in inter-route and inter-modal chains(paths)for possible revisions/changes.展开更多
The primary challenge in weakly supervised semantic segmentation is effectively leveraging weak annotations while minimizing the performance gap compared to fully supervised methods.End-to-end model designs have gaine...The primary challenge in weakly supervised semantic segmentation is effectively leveraging weak annotations while minimizing the performance gap compared to fully supervised methods.End-to-end model designs have gained significant attention for improving training efficiency.Most current algorithms rely on Convolutional Neural Networks(CNNs)for feature extraction.Although CNNs are proficient at capturing local features,they often struggle with global context,leading to incomplete and false Class Activation Mapping(CAM).To address these limitations,this work proposes a Contextual Prototype-Based End-to-End Weakly Supervised Semantic Segmentation(CPEWS)model,which improves feature extraction by utilizing the Vision Transformer(ViT).By incorporating its intermediate feature layers to preserve semantic information,this work introduces the Intermediate Supervised Module(ISM)to supervise the final layer’s output,reducing boundary ambiguity and mitigating issues related to incomplete activation.Additionally,the Contextual Prototype Module(CPM)generates class-specific prototypes,while the proposed Prototype Discrimination Loss and Superclass Suppression Loss guide the network’s training,(LPDL)(LSSL)effectively addressing false activation without the need for extra supervision.The CPEWS model proposed in this paper achieves state-of-the-art performance in end-to-end weakly supervised semantic segmentation without additional supervision.The validation set and test set Mean Intersection over Union(MIoU)of PASCAL VOC 2012 dataset achieved 69.8%and 72.6%,respectively.Compared with ToCo(pre trained weight ImageNet-1k),MIoU on the test set is 2.1%higher.In addition,MIoU reached 41.4%on the validation set of the MS COCO 2014 dataset.展开更多
According to the tensile failure of rock bolt in weakly cemented soft rock, this paper presents a new segmented anchoring style in order to weaken the cumulative effect of anchoring force associated with the large def...According to the tensile failure of rock bolt in weakly cemented soft rock, this paper presents a new segmented anchoring style in order to weaken the cumulative effect of anchoring force associated with the large deformation. Firstly, a segmented mechanical model was established in which free and anchoring section of rock bolt were respectively arranged in different deformation zones. Then, stress and displacement in elastic non-anchoring zone, elastic anchoring zone, elastic sticking zone, softening sticking zone and broken zone were derived respectively based on neural theory and tri-linear strain softening constitutive model of soft rock. Results show that the anchoring effect can be characterized by a supporting parameter b. With its increase, the peak value of tangential stress gradually moves to the roadway wall, and the radial stress significantly increases, which means the decrease of equivalent plastic zone and improvement of confining effect provided by anchorage body. When b increases to 0.72, the equivalent plastic zone disappears, and stresses tend to be the elastic solutions. In addition, the anchoring effect on the displacement of surrounding rock can be quantified by a normalization factor δ.展开更多
As a kind of special material in geotechnical engineering, the mudded weak interlayer plays a crucial part in slope stability. In this paper, we presented a method to determine the threshold value of section micrograp...As a kind of special material in geotechnical engineering, the mudded weak interlayer plays a crucial part in slope stability. In this paper, we presented a method to determine the threshold value of section micrographs of the mudded weak interlayer in slope during its meso-structure qualification process. Some soil tests, scanning electron microscopy（SEM） and image segmentation technology were performed to fulfill our purpose. Specifically, the relation between 3 D-porosity and the threshold was obtained by least square fitting of the threshold-porosity curves and a simplified pore equivalent model. Using this relation and the 3 D-porosity determined by soil experiments, we can figure out the polynomial equation of the threshold value. The threshold values obtained by the other existing methods in literature were employed to validate our present results.展开更多
Weakly supervised semantic segmentation(WSSS)is a tricky task,which only provides category information for segmentation prediction.Thus,the key stage of WSSS is to generate the pseudo labels.For convolutional neural n...Weakly supervised semantic segmentation(WSSS)is a tricky task,which only provides category information for segmentation prediction.Thus,the key stage of WSSS is to generate the pseudo labels.For convolutional neural network(CNN)based methods,in which class activation mapping(CAM)is proposed to obtain the pseudo labels,and only concentrates on the most discriminative parts.Recently,transformer-based methods utilize attention map from the multi-headed self-attention(MHSA)module to predict pseudo labels,which usually contain obvious background noise and incoherent object area.To solve the above problems,we use the Conformer as our backbone,which is a parallel network based on convolutional neural network(CNN)and Transformer.The two branches generate pseudo labels and refine them independently,and can effectively combine the advantages of CNN and Transformer.However,the parallel structure is not close enough in the information communication.Thus,parallel structure can result in poor details about pseudo labels,and the background noise still exists.To alleviate this problem,we propose enhancing convolution CAM(ECCAM)model,which have three improved modules based on enhancing convolution,including deeper stem(DStem),convolutional feed-forward network(CFFN)and feature coupling unit with convolution(FCUConv).The ECCAM could make Conformer have tighter interaction between CNN and Transformer branches.After experimental verification,the improved modules we propose can help the network perceive more local information from images,making the final segmentation results more refined.Compared with similar architecture,our modules greatly improve the semantic segmentation performance and achieve70.2%mean intersection over union(mIoU)on the PASCAL VOC 2012 dataset.展开更多
在图像语义分割领域,无监督领域自适应技术的发展有效降低了模型对标注数据的依赖,提升了自动驾驶等智能系统的效率和广泛适用性。针对无监督领域自适应技术在新场景泛化能力有限及在稀有类别中分割效果差的问题,文章提出了一种基于强...在图像语义分割领域,无监督领域自适应技术的发展有效降低了模型对标注数据的依赖,提升了自动驾驶等智能系统的效率和广泛适用性。针对无监督领域自适应技术在新场景泛化能力有限及在稀有类别中分割效果差的问题,文章提出了一种基于强弱一致性的无监督领域自适应语义分割算法。算法首先通过增加特征级别的增强,拓展图像增强空间的维度,改善了只利用图像级增强局限性。其次,采用基于能量分数的伪标签筛选方法,筛选出足够接近当前训练数据的样本赋予伪标签,避免了使用Softmax置信度方法在稀有类别中存在局限性,使模型更新更加稳健。最后,构建结合图像级别增强和特征级别增强的双重一致性框架,更充分的利用一致性训练,进一步提高模型的泛化能力。实验结果证明,提出的方法在GTA5-to-Cityscapes公开数据集中平均交并比指标(mean Intersection over Union,mIoU)可提升至52.6%,较PixMatch算法,性能提升了4.3%。展开更多
文摘Bus and any other public transit connectivity issues facilitate an understanding of the importance of transit planning in enhancing existing or new transit services.Improving transit connectivity is one of the most vital tasks in transit-operations planning.A poor connection can cause some passengers to stop using the transit service.Service-design criteria always contain postulates to improve routing and scheduling coordination(intra-and inter-agency transfer centers/points and synchronized/timed transfers).Ostensibly the lack of well-defined connectivity measures precludes the weighing and quantifying of the result of any coordination effort.This work provides an initial methodological framework and concepts for(1)quantifying transit connectivity measures and(2)directions and tools for detecting weak segments in inter-route and inter-modal chains(paths)for possible revisions/changes.
基金funding from the following sources:National Natural Science Foundation of China(U1904119)Research Programs of Henan Science and Technology Department(232102210054)+3 种基金Chongqing Natural Science Foundation(CSTB2023NSCQ-MSX0070)Henan Province Key Research and Development Project(231111212000)Aviation Science Foundation(20230001055002)supported by Henan Center for Outstanding Overseas Scientists(GZS2022011).
文摘The primary challenge in weakly supervised semantic segmentation is effectively leveraging weak annotations while minimizing the performance gap compared to fully supervised methods.End-to-end model designs have gained significant attention for improving training efficiency.Most current algorithms rely on Convolutional Neural Networks(CNNs)for feature extraction.Although CNNs are proficient at capturing local features,they often struggle with global context,leading to incomplete and false Class Activation Mapping(CAM).To address these limitations,this work proposes a Contextual Prototype-Based End-to-End Weakly Supervised Semantic Segmentation(CPEWS)model,which improves feature extraction by utilizing the Vision Transformer(ViT).By incorporating its intermediate feature layers to preserve semantic information,this work introduces the Intermediate Supervised Module(ISM)to supervise the final layer’s output,reducing boundary ambiguity and mitigating issues related to incomplete activation.Additionally,the Contextual Prototype Module(CPM)generates class-specific prototypes,while the proposed Prototype Discrimination Loss and Superclass Suppression Loss guide the network’s training,(LPDL)(LSSL)effectively addressing false activation without the need for extra supervision.The CPEWS model proposed in this paper achieves state-of-the-art performance in end-to-end weakly supervised semantic segmentation without additional supervision.The validation set and test set Mean Intersection over Union(MIoU)of PASCAL VOC 2012 dataset achieved 69.8%and 72.6%,respectively.Compared with ToCo(pre trained weight ImageNet-1k),MIoU on the test set is 2.1%higher.In addition,MIoU reached 41.4%on the validation set of the MS COCO 2014 dataset.
基金Financial support for this work was provided by the Scientific Research Foundation of Shandong University of Science and Technology for Recruited Talents of China(No.2015RCJJ042)the National Natural Science Foundation of China(Nos.41472280,51274133)+1 种基金the Promotive Research Fund for Excellent Young and Middle-aged Scientisits of Shandong Province of China(No.BS2015SF005)the Opening Project Fund of Shandong Provincial Key Laboratory of Civil Engineering Disaster Prevention and Mitigation(No.CDPM2013KF05)
文摘According to the tensile failure of rock bolt in weakly cemented soft rock, this paper presents a new segmented anchoring style in order to weaken the cumulative effect of anchoring force associated with the large deformation. Firstly, a segmented mechanical model was established in which free and anchoring section of rock bolt were respectively arranged in different deformation zones. Then, stress and displacement in elastic non-anchoring zone, elastic anchoring zone, elastic sticking zone, softening sticking zone and broken zone were derived respectively based on neural theory and tri-linear strain softening constitutive model of soft rock. Results show that the anchoring effect can be characterized by a supporting parameter b. With its increase, the peak value of tangential stress gradually moves to the roadway wall, and the radial stress significantly increases, which means the decrease of equivalent plastic zone and improvement of confining effect provided by anchorage body. When b increases to 0.72, the equivalent plastic zone disappears, and stresses tend to be the elastic solutions. In addition, the anchoring effect on the displacement of surrounding rock can be quantified by a normalization factor δ.
基金Funded by the National Natural Science Foundation of China(No.51574201)the State Key Laboratory of Geohazard Prevention and Geoenvironment Protection Foundation(Chengdu University of Technology)(No.SKLGP2015K006)the Scientific&Technical Youth Innovation Group(Southwest Petroleum University)(No.2015CXTD05)
文摘As a kind of special material in geotechnical engineering, the mudded weak interlayer plays a crucial part in slope stability. In this paper, we presented a method to determine the threshold value of section micrographs of the mudded weak interlayer in slope during its meso-structure qualification process. Some soil tests, scanning electron microscopy（SEM） and image segmentation technology were performed to fulfill our purpose. Specifically, the relation between 3 D-porosity and the threshold was obtained by least square fitting of the threshold-porosity curves and a simplified pore equivalent model. Using this relation and the 3 D-porosity determined by soil experiments, we can figure out the polynomial equation of the threshold value. The threshold values obtained by the other existing methods in literature were employed to validate our present results.
文摘Weakly supervised semantic segmentation(WSSS)is a tricky task,which only provides category information for segmentation prediction.Thus,the key stage of WSSS is to generate the pseudo labels.For convolutional neural network(CNN)based methods,in which class activation mapping(CAM)is proposed to obtain the pseudo labels,and only concentrates on the most discriminative parts.Recently,transformer-based methods utilize attention map from the multi-headed self-attention(MHSA)module to predict pseudo labels,which usually contain obvious background noise and incoherent object area.To solve the above problems,we use the Conformer as our backbone,which is a parallel network based on convolutional neural network(CNN)and Transformer.The two branches generate pseudo labels and refine them independently,and can effectively combine the advantages of CNN and Transformer.However,the parallel structure is not close enough in the information communication.Thus,parallel structure can result in poor details about pseudo labels,and the background noise still exists.To alleviate this problem,we propose enhancing convolution CAM(ECCAM)model,which have three improved modules based on enhancing convolution,including deeper stem(DStem),convolutional feed-forward network(CFFN)and feature coupling unit with convolution(FCUConv).The ECCAM could make Conformer have tighter interaction between CNN and Transformer branches.After experimental verification,the improved modules we propose can help the network perceive more local information from images,making the final segmentation results more refined.Compared with similar architecture,our modules greatly improve the semantic segmentation performance and achieve70.2%mean intersection over union(mIoU)on the PASCAL VOC 2012 dataset.
文摘在图像语义分割领域,无监督领域自适应技术的发展有效降低了模型对标注数据的依赖,提升了自动驾驶等智能系统的效率和广泛适用性。针对无监督领域自适应技术在新场景泛化能力有限及在稀有类别中分割效果差的问题,文章提出了一种基于强弱一致性的无监督领域自适应语义分割算法。算法首先通过增加特征级别的增强,拓展图像增强空间的维度,改善了只利用图像级增强局限性。其次,采用基于能量分数的伪标签筛选方法,筛选出足够接近当前训练数据的样本赋予伪标签,避免了使用Softmax置信度方法在稀有类别中存在局限性,使模型更新更加稳健。最后,构建结合图像级别增强和特征级别增强的双重一致性框架,更充分的利用一致性训练,进一步提高模型的泛化能力。实验结果证明,提出的方法在GTA5-to-Cityscapes公开数据集中平均交并比指标(mean Intersection over Union,mIoU)可提升至52.6%,较PixMatch算法,性能提升了4.3%。
