Learning a generic 3D face model from 2D image databases using incremental Structure-from-Motion

Over the last decade 3D face models have been extensively used in many applications such as face recognition, facial animation and facial expression analysis. 3D Morphable Models (MMs) have become a popular tool to build and fit 3D face models to images. Critical to the success of MMs is the ability to build a generic 3D face model. Major limitations in the MMs building process are: (1) collecting 3D data usually involves the use of expensive laser scans and complex capture setups, (2) the number of available 3D databases is limited, and typically there is a lack of expression variability and (3) finding correspondences and registering the 3D model is a labor intensive and error prone process.     

The convergence analysis and specification of the Population-Based Incremental Learning algorithm

In this paper, we investigate the global convergence properties in probability of the Population-Based Incremental Learning (PBIL) algorithm when the initial configuration p⁽⁰⁾ is fixed and the learning rate α is close to zero. The convergence in probability of PBIL is confirmed by the experimental results. This paper presents a meaningful discussion on how to establish a unified convergence theory of PBIL that is not affected by the population and the selected individuals.     

一种基于Mahalanobis距离的增量聚类算法

经典的模糊c均值聚类算法对非球型或椭球型分布的数据集进行聚类效果较差。将经典的模糊c均值聚类中的欧氏距离用Mahalanobis距离替代,利用Mahalanobis距离的优点,将其用于增量学习中,提出一种基于马氏距离的模糊增量聚类学习算法。实验结果表明该算法能较有效地解决模糊聚类方法中的缺陷,提高了训练精度。     

Efficient single-pass frequent pattern mining using a prefix-tree

The FP-growth algorithm using the FP-tree has been widely studied for frequent pattern mining because it can dramatically improve performance compared to the candidate generation-and-test paradigm of Apriori. However, it still requires two database scans, which are not consistent with efficient data stream processing. In this paper, we present a novel tree structure, called CP-tree (compact pattern tree), that captures database information with one scan (insertion phase) and provides the same mining performance as the FP-growth method (restructuring phase). The CP-tree introduces the concept of dynamic tree restructuring to produce a highly compact frequency-descending tree structure at runtime. An efficient tree restructuring method, called the branch sorting method, that restructures a prefix-tree branch-by-branch, is also proposed in this paper. Moreover, the CP-tree provides full functionality for interactive and incremental mining. Extensive experimental results show that the CP-tree is efficient for frequent pattern mining, interactive, and incremental mining with a single database scan.     

RMAIN: Association rules maintenance without reruns through data

Association rules are well recognised as a data mining tool for analysis of transactional data, currently going far beyond the early basket-based applications. A wide spectrum of methods for mining associations have been proposed up to date, including batch and incremental approaches. Most of the accurate incremental methods minimise, but do not completely eliminate reruns through processed data. In this paper we propose a new approximate algorithm RMAIN for incremental maintenance of association rules, which works repeatedly on subsequent portions of new transactions. After a portion has been analysed, the new rules are combined with the old ones, so that no reruns through the processed transactions are performed in the future. The resulting set of rules is kept similar to the one that would be achieved in a batch manner. Unlike other incremental methods, RMAIN is fully separated from a rule mining algorithm and this independence makes it highly general and flexible. Moreover, it operates on rules in their final form, ready for decision support, and not on intermediate representation (frequent itemsets), which requires further processing. These features make the RMAIN algorithm well suited for rule maintenance within knowledge bases of autonomous systems with strongly bounded resources and time for decision making. We evaluated the algorithm on synthetic and real datasets, achieving promising results with respect to either performance or quality of output rules.     

Precise Hsu’s method for analyzing the stability of periodic solutions of multi-degrees-of-freedom systems with cubic nonlinearity

This paper presents a new precise Hsu’s method for investigating the stability regions of the periodic motions of an undamped two-degrees-of-freedom system with cubic nonlinearity. Firstly, the incremental harmonic balance (IHB) method is used to obtain the solution of nonlinear vibration differential equations. Hsu’s method is then adopted for computing the transition matrix at the end of one period, and the precise time integration algorithm is adjusted to improve the computational precision. The stability regions of the system obtained from the precise Hsu’s, Hsu’s and improved numerical integration methods are compared and discussed.     

Distance browsing in distributed multimedia databases

The state of the art of searching for non-text data (e.g., images) is to use extracted metadata annotations or text, which might be available as a related information. However, supporting real content-based audiovisual search, based on similarity search on features, is significantly more expensive than searching for text. Moreover, such search exhibits linear scalability with respect to the dataset size, so parallel query execution is needed.In this paper, we present a Distributed Incremental Nearest Neighbor algorithm (DINN) for finding closest objects in an incremental fashion over data distributed among computer nodes, each able to perform its local Incremental Nearest Neighbor (local-INN) algorithm. We prove that our algorithm is optimum with respect to both the number of involved nodes and the number of local-INN invocations. An implementation of our DINN algorithm, on a real P2P system called MCAN, was used for conducting an extensive experimental evaluation on a real-life dataset.The proposed algorithm is being used in two running projects: SAPIR and NeP4B.     

An adaptive genetic-based signature learning system for intrusion detection

Rule-based intrusion detection systems generally rely on hand crafted signatures developed by domain experts. This could lead to a delay in updating the signature bases and potentially compromising the security of protected systems. In this paper, we present a biologically-inspired computational approach to dynamically and adaptively learn signatures for network intrusion detection using a supervised learning classifier system. The classifier is an online and incremental parallel production rule-based system.A signature extraction system is developed that adaptively extracts signatures to the knowledge base as they are discovered by the classifier. The signature extraction algorithm is augmented by introducing new generalisation operators that minimise overlap and conflict between signatures. Mechanisms are provided to adapt main algorithm parameters to deal with online noisy and imbalanced class data. Our approach is hybrid in that signatures for both intrusive and normal behaviours are learnt.The performance of the developed systems is evaluated with a publicly available intrusion detection dataset and results are presented that show the effectiveness of the proposed system.     

在静态编译器中实现Java异常机制的算法

将Java程序静态编译成可执行程序是使用Java虚拟机动态编译／解释执行Java程序的另一种运行Java程序的方式。针对Java异常机制的特点和静态编译的需求,在介绍Java异常处理逻辑的基础上,提出一种在静态编译器中实现Java异常机制的算法,结合Open64开源编译器,给出该算法的具体步骤以及实现方式,以SPECjvm98为测试集,验证该算法的有效性。     

二维SIMD结构的低功耗调度

针对二维SIMD结构,提出一种可以动态关闭空转部件且结合编译器、指令集和体系结构支持的低功耗调度算法,其中包括编译器优化二维SIMD指令,功耗指令发出部件开关信号,系统接收信号并执行。采用对不同功能单元分别调度的方式和部件局部化的方法。在模拟器上的实验结果表明该方法可以节省整个系统约15％的能量消耗。