稀疏卷积在处理激光雷达点云单目标跟踪时的潜力尚未得到充分发掘.目前,绝大多数点云跟踪算法使用基于球邻域的骨干网络,其显存计算资源占用大并且目标感知的关系建模不充分.针对此问题,本文提出一种基于稀疏卷积结构的LiDAR(Lightlaser...稀疏卷积在处理激光雷达点云单目标跟踪时的潜力尚未得到充分发掘.目前,绝大多数点云跟踪算法使用基于球邻域的骨干网络,其显存计算资源占用大并且目标感知的关系建模不充分.针对此问题,本文提出一种基于稀疏卷积结构的LiDAR(Lightlaser Detection And Ranging)点云跟踪算法,并创新性地融合了空间点与体素双通道的关系建模模块,以高效适应稀疏框架下目标判别信息的嵌入.首先,本文采用3D稀疏卷积残差网络来分别提取模板和搜索区域的特征,并利用反卷积来获取逐点特征来保证跟踪任务中对空间位置特性的要求.其次,关系建模模块进一步在模板与搜索区域特征之间计算相似度语义查询表.为了捕捉到模板与搜索区域间细粒度的关联性,该模块一方面在空间点通道中利用近邻算法找出每个搜索区域点的模板近邻点,并根据语义查询表提取对应特征;另一方面,在体素通道中以每个搜索区域点为中心构建局部多尺度体素,并根据落入体素单元的模板点索引计算语义查询表中值的累计和.最后,将双通道的特征融合并送入基于鸟瞰图的候选包围盒生成模块来回归目标包围盒.为了验证所提出方法的优越性,本文在KITTI和NuScenes数据集进行了测试,对比其他使用稀疏卷积的算法,本文方法平均成功率和精确率分别提升了11.0%和12.0%.本文方法在继承了稀疏卷积高效特点的同时还实现了跟踪精度的提高.展开更多
1.Introduction The activation and conversion of the Earth’s abundant inert nitrogen resources into high-value-added products epitomize one of nature’s most critical processes[1-3].Nitric acid(HNO_(3)),as a pivotal i...1.Introduction The activation and conversion of the Earth’s abundant inert nitrogen resources into high-value-added products epitomize one of nature’s most critical processes[1-3].Nitric acid(HNO_(3)),as a pivotal industrial raw material,showcases an extensive range of applications and possesses substantial economic value in the production of indispensable chemical compounds[4-6].展开更多
All-inorganic CsPbIBr_(2) perovskite has attracted widespread attention in photovoltaic and other optoelectronic devices because of its superior thermal stability.However,the deposition of high-quality solutionprocess...All-inorganic CsPbIBr_(2) perovskite has attracted widespread attention in photovoltaic and other optoelectronic devices because of its superior thermal stability.However,the deposition of high-quality solutionprocessed CsPbIBr_(2) perovskite films with large thicknesses remains challenging.Here,we develop a triple-component precursor(TCP) by employing lead bromide,lead iodide,and cesium bromide,to replace the most commonly used double-component precursor(DCP) consisting of lead bromide and cesium iodide.Remarkably,the TCP system significantly increases the solution concentration to 1.3 M,leading to a larger film thickness(~390 nm) and enhanced light absorption.The resultant CsPbIBr_(2) films were evaluated in planar n-i-p structured solar cells,which exhibit a considerably higher optimal photocurrent density of 11.50 mA cm^(-2) in comparison to that of DCP-based devices(10.69 mA cm^(-2)).By adopting an organic surface passivator,the maximum device efficiency using TCP is further boosted to a record efficiency of 12.8% for CsPbIBr_(2) perovskite solar cells.展开更多
Interconnection of all things challenges the traditional communication methods,and Semantic Communication and Computing(SCC)will become new solutions.It is a challenging task to accurately detect,extract,and represent...Interconnection of all things challenges the traditional communication methods,and Semantic Communication and Computing(SCC)will become new solutions.It is a challenging task to accurately detect,extract,and represent semantic information in the research of SCC-based networks.In previous research,researchers usually use convolution to extract the feature information of a graph and perform the corresponding task of node classification.However,the content of semantic information is quite complex.Although graph convolutional neural networks provide an effective solution for node classification tasks,due to their limitations in representing multiple relational patterns and not recognizing and analyzing higher-order local structures,the extracted feature information is subject to varying degrees of loss.Therefore,this paper extends from a single-layer topology network to a multi-layer heterogeneous topology network.The Bidirectional Encoder Representations from Transformers(BERT)training word vector is introduced to extract the semantic features in the network,and the existing graph neural network is improved by combining the higher-order local feature module of the network model representation network.A multi-layer network embedding algorithm on SCC-based networks with motifs is proposed to complete the task of end-to-end node classification.We verify the effectiveness of the algorithm on a real multi-layer heterogeneous network.展开更多
石墨具有成本低、放电稳定等优点,是钾离子电池极具发展前景的负极材料之一,但其倍率性能仍需改进。本文以中间相炭微球(MCMB)为原料,经KOH处理,设计了一种新型的石墨化负极。通过有限的氧化和轻微的嵌入,在MCMB表面形成了层间距增大的...石墨具有成本低、放电稳定等优点,是钾离子电池极具发展前景的负极材料之一,但其倍率性能仍需改进。本文以中间相炭微球(MCMB)为原料,经KOH处理,设计了一种新型的石墨化负极。通过有限的氧化和轻微的嵌入,在MCMB表面形成了层间距增大的膨胀层,K+的扩散系数明显提高。作为负极时,改性MCMB在低于0.25 V下展现出高平台容量(271 mAh g^(-1)),优越的倍率性能(在1.0 A g^(-1)下,容量可达160 mAh g^(-1)),良好的循环稳定性(在0.1 A g^(-1)下循环100圈后,容量维持为184 mAh g^(-1));当采用羧甲基纤维素作为黏结剂时,KOH处理的MCMB具有高的首次库仑效率(79.2%)。本工作为设计具有优良储钾性能的石墨化材料提供了一种简便的策略。展开更多
文摘稀疏卷积在处理激光雷达点云单目标跟踪时的潜力尚未得到充分发掘.目前,绝大多数点云跟踪算法使用基于球邻域的骨干网络,其显存计算资源占用大并且目标感知的关系建模不充分.针对此问题,本文提出一种基于稀疏卷积结构的LiDAR(Lightlaser Detection And Ranging)点云跟踪算法,并创新性地融合了空间点与体素双通道的关系建模模块,以高效适应稀疏框架下目标判别信息的嵌入.首先,本文采用3D稀疏卷积残差网络来分别提取模板和搜索区域的特征,并利用反卷积来获取逐点特征来保证跟踪任务中对空间位置特性的要求.其次,关系建模模块进一步在模板与搜索区域特征之间计算相似度语义查询表.为了捕捉到模板与搜索区域间细粒度的关联性,该模块一方面在空间点通道中利用近邻算法找出每个搜索区域点的模板近邻点,并根据语义查询表提取对应特征;另一方面,在体素通道中以每个搜索区域点为中心构建局部多尺度体素,并根据落入体素单元的模板点索引计算语义查询表中值的累计和.最后,将双通道的特征融合并送入基于鸟瞰图的候选包围盒生成模块来回归目标包围盒.为了验证所提出方法的优越性,本文在KITTI和NuScenes数据集进行了测试,对比其他使用稀疏卷积的算法,本文方法平均成功率和精确率分别提升了11.0%和12.0%.本文方法在继承了稀疏卷积高效特点的同时还实现了跟踪精度的提高.
基金supported by the National Natural Science Foundation of China,China(22109078,21908120 and 22206094)the Natural Science Outstanding Youth Fund of Shandong Province,China(ZR2023YQ018)the Youth Natural Science Foundation of Hunan Province,China(2021JJ540044)。
文摘1.Introduction The activation and conversion of the Earth’s abundant inert nitrogen resources into high-value-added products epitomize one of nature’s most critical processes[1-3].Nitric acid(HNO_(3)),as a pivotal industrial raw material,showcases an extensive range of applications and possesses substantial economic value in the production of indispensable chemical compounds[4-6].
基金The authors acknowledge the financial support by the National Natural Science Foundation of China(52161145408 and 21975038)the Research and Innovation Team Project of Dalian University of Technology(DUT2022TB10)+2 种基金the Fundamental Research Funds for the Central Universities(DUT22QN213)the Innovation Technology Fund(MRP/040/21X)the Green Technology Fund(GTF202020164)for their financial support。
文摘All-inorganic CsPbIBr_(2) perovskite has attracted widespread attention in photovoltaic and other optoelectronic devices because of its superior thermal stability.However,the deposition of high-quality solutionprocessed CsPbIBr_(2) perovskite films with large thicknesses remains challenging.Here,we develop a triple-component precursor(TCP) by employing lead bromide,lead iodide,and cesium bromide,to replace the most commonly used double-component precursor(DCP) consisting of lead bromide and cesium iodide.Remarkably,the TCP system significantly increases the solution concentration to 1.3 M,leading to a larger film thickness(~390 nm) and enhanced light absorption.The resultant CsPbIBr_(2) films were evaluated in planar n-i-p structured solar cells,which exhibit a considerably higher optimal photocurrent density of 11.50 mA cm^(-2) in comparison to that of DCP-based devices(10.69 mA cm^(-2)).By adopting an organic surface passivator,the maximum device efficiency using TCP is further boosted to a record efficiency of 12.8% for CsPbIBr_(2) perovskite solar cells.
基金supported by National Natural Science Foundation of China(62101088,61801076,61971336)Natural Science Foundation of Liaoning Province(2022-MS-157,2023-MS-108)+1 种基金Key Laboratory of Big Data Intelligent Computing Funds for Chongqing University of Posts and Telecommunications(BDIC-2023-A-003)Fundamental Research Funds for the Central Universities(3132022230).
文摘Interconnection of all things challenges the traditional communication methods,and Semantic Communication and Computing(SCC)will become new solutions.It is a challenging task to accurately detect,extract,and represent semantic information in the research of SCC-based networks.In previous research,researchers usually use convolution to extract the feature information of a graph and perform the corresponding task of node classification.However,the content of semantic information is quite complex.Although graph convolutional neural networks provide an effective solution for node classification tasks,due to their limitations in representing multiple relational patterns and not recognizing and analyzing higher-order local structures,the extracted feature information is subject to varying degrees of loss.Therefore,this paper extends from a single-layer topology network to a multi-layer heterogeneous topology network.The Bidirectional Encoder Representations from Transformers(BERT)training word vector is introduced to extract the semantic features in the network,and the existing graph neural network is improved by combining the higher-order local feature module of the network model representation network.A multi-layer network embedding algorithm on SCC-based networks with motifs is proposed to complete the task of end-to-end node classification.We verify the effectiveness of the algorithm on a real multi-layer heterogeneous network.
文摘石墨具有成本低、放电稳定等优点,是钾离子电池极具发展前景的负极材料之一,但其倍率性能仍需改进。本文以中间相炭微球(MCMB)为原料,经KOH处理,设计了一种新型的石墨化负极。通过有限的氧化和轻微的嵌入,在MCMB表面形成了层间距增大的膨胀层,K+的扩散系数明显提高。作为负极时,改性MCMB在低于0.25 V下展现出高平台容量(271 mAh g^(-1)),优越的倍率性能(在1.0 A g^(-1)下,容量可达160 mAh g^(-1)),良好的循环稳定性(在0.1 A g^(-1)下循环100圈后,容量维持为184 mAh g^(-1));当采用羧甲基纤维素作为黏结剂时,KOH处理的MCMB具有高的首次库仑效率(79.2%)。本工作为设计具有优良储钾性能的石墨化材料提供了一种简便的策略。