Few-shot Learning algorithms can be effectively applied to fields where certain categories have only a small amount of data or a small amount of labeled data,such as medical images,terrorist surveillance,and so on.The...Few-shot Learning algorithms can be effectively applied to fields where certain categories have only a small amount of data or a small amount of labeled data,such as medical images,terrorist surveillance,and so on.The Metric Learning in the Few-shot Learning algorithmis classified by measuring the similarity between the classified samples and the unclassified samples.This paper improves the Prototypical Network in the Metric Learning,and changes its core metric function to Manhattan distance.The Convolutional Neural Network of the embedded module is changed,and mechanisms such as average pooling and Dropout are added.Through comparative experiments,it is found that thismodel can converge in a small number of iterations(below 15,000 episodes),and its performance exceeds algorithms such asMAML.Research shows that replacingManhattan distance with Euclidean distance can effectively improve the classification effect of the Prototypical Network,and mechanisms such as average pooling and Dropout can also effectively improve the model.展开更多
A novel framework for non-Manhattan channel routing considering crosstalk and wire length reduction is proposed. To handle crosstalk and wire length reduction problem, an improved non-Manhattan router in the gridded r...A novel framework for non-Manhattan channel routing considering crosstalk and wire length reduction is proposed. To handle crosstalk and wire length reduction problem, an improved non-Manhattan router in the gridded routing model is developed, and a novel rerouting algorithm based on segment-rerouting technique and layer reassignment is presented. Simulation experiments demonstrate that our algorithm achieves very good results.展开更多
针对传统的ORB(Oriented fast and rotated brief)算法在运算速度以及精度方面有时难以满足某些应用场合实际要求,在特征点提取阶段,利用金字塔光流法提取特征点并划分有效及无效区域特征点,从而降低特征点匹配个数和提高后续运算特征...针对传统的ORB(Oriented fast and rotated brief)算法在运算速度以及精度方面有时难以满足某些应用场合实际要求,在特征点提取阶段,利用金字塔光流法提取特征点并划分有效及无效区域特征点,从而降低特征点匹配个数和提高后续运算特征点匹配速度;在特征点匹配阶段,将传统算法中的欧氏距离改为曼哈顿距离,再用MLESAC算法来剔除误匹配点。将SURF(Speeded up robust features)算法、SIFT(Scale-invariant feature transform)算法、ORB算法和改进后的ORB算法对光照条件不同、模糊度不同以及尺度大小不同的两张图像进行处理,改进后的ORB算法无论是在匹配速度还是匹配精度方面相比于传统ORB算法都有了明显改善。展开更多
为提高ORB-SLAM2 (oriented fast and rotated brief, and simultaneous localization and mapping)系统的位姿估计精度并解决仅能生成稀疏地图的问题,提出了融合迭代最近点拟合(iterative closest point, ICP)算法与曼哈顿世界假说的...为提高ORB-SLAM2 (oriented fast and rotated brief, and simultaneous localization and mapping)系统的位姿估计精度并解决仅能生成稀疏地图的问题,提出了融合迭代最近点拟合(iterative closest point, ICP)算法与曼哈顿世界假说的位姿估计策略并在系统中加入稠密建图线程。首先通过ORB(oriented fast and rotated brief)特征点法、最小显著性差异(least-significant difference, LSD)算法和聚集层次聚类(agglomerative hierarchical clustering, AHC)方法提取点、线、面特征,其中点、线特征与上一帧匹配,面特征在全局地图匹配。然后采用基于surfel的稠密建图策略将图像划分为非平面与平面区域,非平面采用ICP算法计算位姿,平面则通过面与面的正交关系确定曼哈顿世界从而使用不同估计策略,其中曼哈顿世界场景通过位姿解耦实现基于曼哈顿帧观测的无漂移旋转估计,而该场景的平移以及非曼哈顿世界场景的位姿采用追踪的点、线、面特征进行估计和优化;最后根据关键帧和相应位姿实现稠密建图。采用慕尼黑工业大学(technische universit?t münchen, TUM)数据集验证所提建图方法,经过与ORB-SLAM2算法比较,均方根误差平均减少0.24 cm,平均定位精度提高7.17%,验证了所提方法进行稠密建图的可行性和有效性。展开更多
We propose an approach for automatic generation of building models by assembling a set of boxes using a Manhattan-world assumption.The method first aligns the point cloud with a per-building local coordinate system,an...We propose an approach for automatic generation of building models by assembling a set of boxes using a Manhattan-world assumption.The method first aligns the point cloud with a per-building local coordinate system,and then fits axis-aligned planes to the point cloud through an iterative regularization process.The refined planes partition the space of the data into a series of compact cubic cells(candidate boxes)spanning the entire 3D space of the input data.We then choose to approximate the target building by the assembly of a subset of these candidate boxes using a binary linear programming formulation.The objective function is designed to maximize the point cloud coverage and the compactness of the final model.Finally,all selected boxes are merged into a lightweight polygonal mesh model,which is suitable for interactive visualization of large scale urban scenes.Experimental results and a comparison with state-of-the-art methods demonstrate the effectiveness of the proposed framework.展开更多
This paper critically reexamines the canonical utopic scene from the end of Johann Wolfgang von Goethe's Faust II(1832)through the writings of the British-born Yale physicist Dr.William Francis Gray Swann(1884-196...This paper critically reexamines the canonical utopic scene from the end of Johann Wolfgang von Goethe's Faust II(1832)through the writings of the British-born Yale physicist Dr.William Francis Gray Swann(1884-1962),who composed one hundred years later a book on the philosophy of contemporary physical science titled The Architecture o f the Universe(1934).In his text,Swann invokes the instrumentalization of atomic energy in a bargain to defeat a cosmic anxiety borne from the modem temporal experience of entropy,a pact whose culmination is akin to Faust's seminal turning away of the ghost Sorge(Care),with strong ramifications for the future postatomic world order.The physical initiation of this“zero hour”is best known through the work of Swann’s Ph.D.student,Ernest Orlando Lawrence,who went on to pioneer the mass separation of electromagnetic isotopes at Oak Ridge.This paper will posit the Manhattan Project as part of the legacy of a pre-war universal imaginary,in order to introduce a new critical reading o f utopia in Germany's greatest poet.展开更多
文摘Few-shot Learning algorithms can be effectively applied to fields where certain categories have only a small amount of data or a small amount of labeled data,such as medical images,terrorist surveillance,and so on.The Metric Learning in the Few-shot Learning algorithmis classified by measuring the similarity between the classified samples and the unclassified samples.This paper improves the Prototypical Network in the Metric Learning,and changes its core metric function to Manhattan distance.The Convolutional Neural Network of the embedded module is changed,and mechanisms such as average pooling and Dropout are added.Through comparative experiments,it is found that thismodel can converge in a small number of iterations(below 15,000 episodes),and its performance exceeds algorithms such asMAML.Research shows that replacingManhattan distance with Euclidean distance can effectively improve the classification effect of the Prototypical Network,and mechanisms such as average pooling and Dropout can also effectively improve the model.
文摘A novel framework for non-Manhattan channel routing considering crosstalk and wire length reduction is proposed. To handle crosstalk and wire length reduction problem, an improved non-Manhattan router in the gridded routing model is developed, and a novel rerouting algorithm based on segment-rerouting technique and layer reassignment is presented. Simulation experiments demonstrate that our algorithm achieves very good results.
文摘针对传统的ORB(Oriented fast and rotated brief)算法在运算速度以及精度方面有时难以满足某些应用场合实际要求,在特征点提取阶段,利用金字塔光流法提取特征点并划分有效及无效区域特征点,从而降低特征点匹配个数和提高后续运算特征点匹配速度;在特征点匹配阶段,将传统算法中的欧氏距离改为曼哈顿距离,再用MLESAC算法来剔除误匹配点。将SURF(Speeded up robust features)算法、SIFT(Scale-invariant feature transform)算法、ORB算法和改进后的ORB算法对光照条件不同、模糊度不同以及尺度大小不同的两张图像进行处理,改进后的ORB算法无论是在匹配速度还是匹配精度方面相比于传统ORB算法都有了明显改善。
文摘为提高ORB-SLAM2 (oriented fast and rotated brief, and simultaneous localization and mapping)系统的位姿估计精度并解决仅能生成稀疏地图的问题,提出了融合迭代最近点拟合(iterative closest point, ICP)算法与曼哈顿世界假说的位姿估计策略并在系统中加入稠密建图线程。首先通过ORB(oriented fast and rotated brief)特征点法、最小显著性差异(least-significant difference, LSD)算法和聚集层次聚类(agglomerative hierarchical clustering, AHC)方法提取点、线、面特征,其中点、线特征与上一帧匹配,面特征在全局地图匹配。然后采用基于surfel的稠密建图策略将图像划分为非平面与平面区域,非平面采用ICP算法计算位姿,平面则通过面与面的正交关系确定曼哈顿世界从而使用不同估计策略,其中曼哈顿世界场景通过位姿解耦实现基于曼哈顿帧观测的无漂移旋转估计,而该场景的平移以及非曼哈顿世界场景的位姿采用追踪的点、线、面特征进行估计和优化;最后根据关键帧和相应位姿实现稠密建图。采用慕尼黑工业大学(technische universit?t münchen, TUM)数据集验证所提建图方法,经过与ORB-SLAM2算法比较,均方根误差平均减少0.24 cm,平均定位精度提高7.17%,验证了所提方法进行稠密建图的可行性和有效性。
文摘We propose an approach for automatic generation of building models by assembling a set of boxes using a Manhattan-world assumption.The method first aligns the point cloud with a per-building local coordinate system,and then fits axis-aligned planes to the point cloud through an iterative regularization process.The refined planes partition the space of the data into a series of compact cubic cells(candidate boxes)spanning the entire 3D space of the input data.We then choose to approximate the target building by the assembly of a subset of these candidate boxes using a binary linear programming formulation.The objective function is designed to maximize the point cloud coverage and the compactness of the final model.Finally,all selected boxes are merged into a lightweight polygonal mesh model,which is suitable for interactive visualization of large scale urban scenes.Experimental results and a comparison with state-of-the-art methods demonstrate the effectiveness of the proposed framework.
文摘This paper critically reexamines the canonical utopic scene from the end of Johann Wolfgang von Goethe's Faust II(1832)through the writings of the British-born Yale physicist Dr.William Francis Gray Swann(1884-1962),who composed one hundred years later a book on the philosophy of contemporary physical science titled The Architecture o f the Universe(1934).In his text,Swann invokes the instrumentalization of atomic energy in a bargain to defeat a cosmic anxiety borne from the modem temporal experience of entropy,a pact whose culmination is akin to Faust's seminal turning away of the ghost Sorge(Care),with strong ramifications for the future postatomic world order.The physical initiation of this“zero hour”is best known through the work of Swann’s Ph.D.student,Ernest Orlando Lawrence,who went on to pioneer the mass separation of electromagnetic isotopes at Oak Ridge.This paper will posit the Manhattan Project as part of the legacy of a pre-war universal imaginary,in order to introduce a new critical reading o f utopia in Germany's greatest poet.
