标记判别和局部线性强化的半监督稀疏子空间聚类

子空间聚类通常可以很好地处理高维数据,但由于数据本身的噪声等的影响,系数矩阵的块对角线结构往往容易被破坏.针对上述问题,提出了一种标记判别和局部线性强化的半监督稀疏子空间聚类.一方面,通过约束标记数据之间的系数为0,更好地捕获数据的全局结构;另一方面,通过K近邻关系加强数据邻近点之间的局部相关性,同时消除大量不相关的数据点,增强算法的鲁棒性.通过在多种数据上的实验,验证了提出的半监督聚类算法的有效性.     

邻域参数动态变化的局部线性嵌入

局部线性嵌入是最有竞争力的非线性降维方法,有较强的表达能力和计算优势.但它们都采用全局一致的邻城大小,只适用于均匀分布的流形,无法处理现实中大量存在的非均匀分布流形.为此,提出一种邻域大小动态确定的新局部线性嵌入方法.它采用Hessian局部线性嵌入的概念框架,但用每个点的局部邻域估计此邻域内任意点之间的近似测地距离,然后根据近似测地距离与欧氏距离之间的关系动态确定该点的邻域大小,并以此邻域大小构造新的局部邻域.算法几何意义清晰,在观察数据稀疏和数据带噪音等情况下,都比现有算法有更强的鲁棒性.标准数据集上的实验结果验证了所提方法的有效性.     

离散点云原始形状及边界曲线提取算法

大规模离散点云包含多种类型的扫描缺陷:噪声、异常数据、孔洞及不规则的各向异性采样,大部分现有的算法不能够很好地处理这些缺陷,这对点云拓扑关系的恢复及特征提取带来了困难.针对此问题,提出了一种健壮有效的点云重构算法,首先,计算每个数据点的局部属性;然后利用局部属性探测点云中包含的原始形状;最后利用统计优化方法对原始形状中...     

散乱点云数据曲率估计方法

针对带有强噪声离散点云数据曲率计算问题,提出一种基于稳健统计的曲率估计方法。首先,用一个二次曲面拟合三维空间采样点处的局部形状;其次,随机地选择该采样点邻域内的子集,多次执行这样的拟合过程,通过变窗宽的最大核密度估计,就得到了最优拟合曲面;最后,将采样点投影到该曲面上,计算投影点曲率信息,就得到采样点曲率。实验结果表明,所提方法对噪声和离群点是稳健的,特别是随着噪声方差的增大,要明显好于传统的抛物拟合方法。     

点云模型基于几何特征增强与保特征曲面拟合的法向估计方法EI北大核心CSCD

现有法向估计中,尤其是模型中存在较大噪声的情况下,目标点邻域的选择是一个关键且困难的问题.针对点云模型,为了提高法向估计准确度,提出一种自适应选择邻域且保特征、抗噪声的法向估计方法.首先,提出双边非局部特征增强模块,根据网络前置学习特征以及邻域点几何特性对点邻域进行加权选择,并据此对局部特征进行增强,以提升网络对模型局部几何特征学习的能力;然后,采用局部特征与全局特征相结合的形式刻画点云完备的几何特征,并以此为基础进行局部曲面拟合及法向估计;最后,在局部曲面拟合中提出邻域保特征损失,依据邻域点受噪声干扰度对邻域点的拟合权重进行调整,实现保细节特征的局部曲面拟合,提高对噪声的鲁棒性.实验使用PCPNET数据集进行模型训练和测试,大量定性与定量的实验结果表明,与相关方法相比,所提方法对于不同噪声级别以及不同密度分布等复杂情形都可取得更加准确的法向估计结果,并更好地推动曲面重建等点云处理应用.     

一种基于RANSAC的点云特征线提取算法

点云中提取的特征线在点云处理中具有重要的应用价值,已被应用于对称性检测、表面重建及点云与图像之间的注册等。然而,已有的点云特征线提取算法无法有效地处理点云中不可避免的噪声、外点和数据缺失,而随机采样一致性RANSAC由于具有较高的鲁棒性,在图像和三维模型处理中具有广泛的应用。为此,针对由建筑物或机械部件等具有平面特征的物体扫描得到的点云,提出了一种基于RANSAC的特征线提取算法。本算法首先基于RANSAC在点云中检测出多个平面,然后将每个平面参数化域的边界点作为候选,在这些候选点上再应用基于全局约束的RANSAC得到最终的特征线。实验结果表明,该算法对点云中的噪声、外点和数据缺失具有很强的鲁棒性。     

基于多尺度核函数的散乱点云数据过滤方法*

针对光滑曲面采样散乱点云含有噪声及异常数据的问题,提出了一种基于多尺度核函数的过滤处理方法。采用核密度估计技术及均值漂移跟踪算法对原始点云数据进行聚类,结合局部似然函数来测度一个三维点位于采样曲面上的概率,利用过滤后的极大似然点集精确地逼近采样曲面,最后结合经典网格化算法能够获得较好的曲面重构效果。处理实例证明,该方法实用性好,不仅能够很好地抑制不同幅值的噪声,同时也能够探测到异常数据并进行自动清除。     

一种基于数据垂直划分的分布式密度聚类算法

聚类分析是数据挖掘领域的一项重要研究课题,对大数据集的聚类更以其数据量大、噪声数据多等而成为一个难点.针对数据垂直划分的情况,提出连通点集及局部噪声点集等概念.在分析局部噪声点集与全局噪声点集以及局部连通点集与全局连通点集关系的基础上,对全局噪声点进行有效过滤,进一步设计闭三角链表结构存储各个结点的聚类中间结果,提出了基于密度的分布式聚类算法DDBSCAN.理论分析和实验结果表明,算法可以有效解决垂直划分的大数据集聚类问题,算法是有效可行的.     

融合全局和局部特征的下一个兴趣点推荐方法

随着海量移动数据的积累,下一个兴趣点推荐已成为基于位置的社交网络中的一项重要任务.目前,主流方法倾向于从用户近期的签到序列中捕捉局部动态偏好,但忽略了历史移动数据蕴含的全局静态信息,从而阻碍了对用户偏好的进一步挖掘,影响了推荐的准确性.为此,提出一种基于全局和局部特征融合的下一个兴趣点推荐方法.该方法利用签到序列中的顺序依赖和全局静态信息中用户与兴趣点之间、连续签到之间隐藏的关联关系建模用户移动行为.首先,引入两类全局静态信息,即User-POI关联路径和POI-POI关联路径,学习用户的全局静态偏好和连续签到之间的全局依赖关系.具体地,利用交互数据以及地理信息构建异构信息网络,设计关联关系表示学习方法,利用相关度引导的路径采样策略以及层级注意力机制获取全局静态特征.然后,基于两类全局静态特征更新签到序列中的兴趣点表示,并采用位置与时间间隔感知的自注意力机制来捕捉用户签到序列中签到之间的局部顺序依赖,进而评估用户访问兴趣点概率,实现下一个兴趣点推荐.最后,在两个真实数据集上进行了实验比较与分析,验证了所提方法能够有效提升下一个兴趣点推荐的准确性.此外,案例分析表明,建模显式路径有助于提...     

全局结构引导的人造物体参数化基元检测

点云是表达三维数据的常见形式,点云数据提取出的几何基元能够帮助人们快速地理解并处理场景信息,也方便后续其他任务的开展.为了更好地利用人造物体中普遍存在的全局结构关系,增强基元检测过程中全局结构的正向引导,提出参数化基元检测网络——RelationNet,包括2个子模块.首先,为了更好地编码三维点与其所在基元的结构关系,通过空间偏移预测模块预测三维点所在基元中心的偏移向量,提升点对其所在基元的位置感知能力,为后续分割任务提供更多的特征依据;其次,人造物体的基元与基元之间常常具有如平行、垂直、轴对齐等结构关系,为了更好地利用这些关系实现对几何基元检测结果的改进,还包含全局结构关系提取模块,利用基元拟合后获得的参数判断各个基元之间的结构关系,并通过设置相应的损失函数对提取到的结果进行引导监督.在大型ABC数据集与基元监督拟合(SPFN), ParseNet等主流算法进行对比的实验结果表明, RelationNet在基元分割和基元分类任务上的MIoU分别达到85.32%和90.10%,与当前先进方法相比有明显的效果提升.     

Locally linear discriminant analysis for multimodally distributed classes for face recognition with a single model image

We present a novel method of nonlinear discriminant analysis involving a set of locally linear transformations called "Locally Linear Discriminant Analysis" (LLDA). The underlying idea is that global nonlinear data structures are locally linear and local structures can be linearly aligned. Input vectors are projected into each local feature space by linear transformations found to yield locally linearly transformed classes that maximize the between-class covariance while minimizing the within-class covariance. In face recognition, linear discriminant analysis (LIDA) has been widely adopted owing to its efficiency, but it does not capture nonlinear manifolds of faces which exhibit pose variations. Conventional nonlinear classification methods based on kernels such as generalized discriminant analysis (GDA) and support vector machine (SVM) have been developed to overcome the shortcomings of the linear method, but they have the drawback of high computational cost of classification and overfitting. Our method is for multiclass nonlinear discrimination and it is computationally highly efficient as compared to GDA. The method does not suffer from overfitting by virtue of the linear base structure of the solution. A novel gradient-based learning algorithm is proposed for finding the optimal set of local linear bases. The optimization does not exhibit a local-maxima problem. The transformation functions facilitate robust face recognition in a low-dimensional subspace, under pose variations, using a single model image. The classification results are given for both synthetic and real face data.     

Decentralized multi-dimensional alert correlation for collaborative intrusion detection

The growth in coordinated network attacks such as scans, worms and distributed denial-of-service (DDoS) attacks is a profound threat to the security of the Internet. Collaborative intrusion detection systems (CIDSs) have the potential to detect these attacks, by enabling all the participating intrusion detection systems (IDSs) to share suspicious intelligence with each other to form a global view of the current security threats. Current correlation algorithms in CIDSs are either too simple to capture the important characteristics of attacks, or too computationally expensive to detect attacks in a timely manner. We propose a decentralized, multi-dimensional alert correlation algorithm for CIDSs to address these challenges. A multi-dimensional alert clustering algorithm is used to extract the significant intrusion patterns from raw intrusion alerts. A two-stage correlation algorithm is used, which first clusters alerts locally at each IDS, before reporting significant alert patterns to a global correlation stage. We introduce a probabilistic approach to decide when a pattern at the local stage is sufficiently significant to warrant correlation at the global stage. We then implement the proposed two-stage correlation algorithm in a fully distributed CIDS. Our experiments on a large real-world intrusion data set show that our approach can achieve a significant reduction in the number of alert messages generated by the local correlation stage with negligible false negatives compared to a centralized scheme. The proposed probabilistic threshold approach gains a significant improvement in detection accuracy in a stealthy attack scenario, compared to a naive scheme that uses the same threshold at the local and global stages. A large scale experiment on PlanetLab shows that our decentralized architecture is significantly more efficient than a centralized approach in terms of the time required to correlate alerts.     

Spatial Data Mining: Database Primitives, Algorithms and Efficient DBMS Support

Spatial data mining algorithms heavily depend on the efficient processing of neighborhood relations since the neighbors of many objects have to be investigated in a single run of a typical algorithm. Therefore, providing general concepts for neighborhood relations as well as an efficient implementation of these concepts will allow a tight integration of spatial data mining algorithms with a spatial database management system. This will speed up both, the development and the execution of spatial data mining algorithms. In this paper, we define neighborhood graphs and paths and a small set of database primitives for their manipulation. We show that typical spatial data mining algorithms are well supported by the proposed basic operations. For finding significant spatial patterns, only certain classes of paths “leading away” from a starting object are relevant. We discuss filters allowing only such neighborhood paths which will significantly reduce the search space for spatial data mining algorithms. Furthermore, we introduce neighborhood indices to speed up the processing of our database primitives. We implemented the database primitives on top of a commercial spatial database management system. The effectiveness and efficiency of the proposed approach was evaluated by using an analytical cost model and an extensive experimental study on a geographic database.     

Splat-based surface reconstruction from defect-laden point sets

We introduce a method for surface reconstruction from point sets that is able to cope with noise and outliers. First, a splat-based representation is computed from the point set. A robust local 3D RANSAC-based procedure is used to filter the point set for outliers, then a local jet surface – a low-degree surface approximation – is fitted to the inliers. Second, we extract the reconstructed surface in the form of a surface triangle mesh through Delaunay refinement. The Delaunay refinement meshing approach requires computing intersections between line segment queries and the surface to be meshed. In the present case, intersection queries are solved from the set of splats through a 1D RANSAC procedure.     

Teacher feedback to scaffold and refine demonstrated motion primitives on a mobile robot

Task demonstration is an effective technique for developing robot motion control policies. As tasks become more complex, however, demonstration can become more difficult. In this work, we introduce an algorithm that uses corrective human feedback to build a policy able to perform a novel task, by combining simpler policies learned from demonstration. While some demonstration-based learning approaches do adapt policies with execution experience, few provide corrections within low-level motion control domains or to enable the linking of multiple of demonstrated policies. Here we introduce Feedback for Policy Scaffolding (FPS) as an algorithm that first evaluates and corrects the execution of motion primitive policies learned from demonstration. The algorithm next corrects and enables the execution of a more complex task constructed from these primitives. Key advantages of building a policy from demonstrated primitives is the potential for primitive policy reuse within multiple complex policies and the faster development of these policies, in addition to the development of complex policies for which full demonstration is difficult. Policy reuse under our algorithm is assisted by human teacher feedback, which also contributes to the improvement of policy performance. Within a simulated robot motion control domain we validate that, using FPS, a policy for a novel task is successfully built from motion primitives learned from demonstration. We show feedback to both aid and enable policy development, improving policy performance in success, speed and efficiency.     

Alignment of overlapping locally scaled patches for multidimensional scaling and dimensionality reduction

Data observations that lie on a manifold can be approximated by a collection of overlapping local patches, the alignment of which in a low dimensional Euclidean space provides an embedding of the data. This paper describes an embedding method using classical multidimensional scaling as a local model based on the fact that a manifold locally resembles an Euclidean space. A set of overlapping neighborhoods are chosen by a greedy approximation algorithm of minimum set cover. Local patches derived from the set of overlapping neighborhoods by classical multidimensional scaling are aligned in order to minimize a residual measure, which has a quadratic form of the resulting global coordinates and can be minimized analytically by solving an eigenvalue problem. This method requires only distances within each neighborhood and provides locally isometric embedding results. The size of the eigenvalue problem scales with the number of overlapping neighborhoods rather than the number of data points. Experiments on both synthetic and real world data sets demonstrate the effectiveness of this method. Extensions and variations of the method are discussed.     

Supervised learning of large perceptual organization: graphspectral partitioning and learning automata

Perceptual organization offers an elegant framework to group low-level features that are likely to come from a single object. We offer a novel strategy to adapt this grouping process to objects in a domain. Given a set of training images of objects in context, the associated learning process decides on the relative importance of the basic salient relationships such as proximity, parallelness, continuity, junctions, and common region toward segregating the objects from the background. The parameters of the grouping process are cast as probabilistic specifications of Bayesian networks that need to be learned. This learning is accomplished using a team of stochastic automata in an N-player cooperative game framework. The grouping process, which is based on graph partitioning is able to form large groups from relationships defined over a small set of primitives and is fast. We statistically demonstrate the robust performance of the grouping and the learning frameworks on a variety of real images. Among the interesting conclusions is the significant role of photometric attributes in grouping and the ability to form large salient groups from a set of local relations, each defined over a small number of primitives     

Rewriting with Equivalence Relations in ACL2

Traditionally, a conditional rewrite rule directs replacement of one term by another term that is provably equal to it, perhaps under some hypotheses. This paper generalizes the notion of rewrite rule to permit the connecting relation to be merely an equivalence relation. We then extend the algorithm for applying rewrite rules. Applications of these generalized rewrite rules are only admissible in certain equivalential contexts, so the algorithm tracks which equivalence relations are to be preserved and admissible generalized rewrite rules are selected according to this context. We introduce the notions of congruence rule and refinement rule. We also introduce the idea of generated equivalences, corresponding to a new equivalence relation generated by a set of pre-existing ones. Generated equivalences are used to give the rewriter broad access to admissible generalized rewrite rules. We discuss the implementation of these notions in the ACL2 theorem prover. However, the discussion does not assume familiarity with ACL2, and these ideas can be applied to other reasoning systems as well.     

Distribution of a simple shared dataspace architecture

We study a simple software architecture, in which components are coordinated by writing into and reading from a global set. This simple architecture is inspired by the industrial software architecture Splice. We present two results. First, a distributed implementation of the architecture is given and proved correct formally. In the implementation, local sets are maintained and data items are exchanged between these local sets. Next we show that the architecture is sufficiently expressive in principle. In particular, every global specification of a system's behaviour can be divided into components, which coordinate by read and write primitives on a global set only. We heavily rely on recent concepts and proof methods from process algebra.     

Segment Tracing Using Local Lipschitz Bounds

We introduce Segment Tracing, a new algorithm that accelerates the classical Sphere Tracing method for computing the intersection between a ray and an implicit surface. Our approach consists in computing the Lipschitz bound locally over a segment to improve the marching step computation and accelerate the overall process. We describe the computation of the Lipschitz bound for different operators and primitives. We demonstrate that our algorithm significantly reduces the number of field function queries compared to previous methods, without the need for additional accelerating data-structures. Our method can be applied to a vast variety of implicit models ranging from hierarchical procedural objects built from complex primitives, to simulation-generated implicit surfaces created from many particles.