访存敏感的增量式MPSoC应用映射

现代多处理器片上系统(multiprocessor system-on-chip, MPSoC)通常采用片上网络(network-on-chip, NoC)作为其基本互连结构,应用映射是基于片上网络互连的MPSoC设计中的关键问题,应用映射决定应用划分成的各个任务到片上网络节点的分配.许多基于片上网络互连的MPSoC系统将共享存储作为网络中的独立节点,针对这类MPSoC系统,提出一种访存敏感的增量式动态映射策略.该策略离线分析获取应用的访存特征,运行中当应用到达系统时,根据其访存特征选择不同的映射算法,将热点应用围绕共享存储器布局,非热点应用远离共享存储器布局,并最小化应用间以及应用所含任务间的通信链路竞争.模拟实验表明：与贪恋区域选择加随机节点映射的策略相比较,提出的策略对系统整体通信功耗平均节约34.6%,性能提升可达36.3%,并能适应不同片上网络规模.     

人脸合成中模型的平滑调整和逼真的纹理映射

将薄平面样条函数用于人头模型的调整,使模型调整变得更加平滑、自然,并在纹理映射时,提出了基于模型分块的视角相关纹理映射的方法,该方法将模型分块并用每块对应的最清晰、非冗余的纹理对模型块做映射,使块与块之间形成的规则过渡带非常易于纹理融合,整个模型逐块地一次绘制完成,然后在模型中嵌入了抽象的人脸肌肉,合成出非常逼真的人脸表情动画．实验表明文中算法快速、简捷,合成的人脸图像细节清晰、逼真．     

Energy Efficient Run-Time Incremental Mapping for 3-D Networks-on-Chip

3-D Networks-on-Chip(NoC) emerge as a potent solution to address both the interconnection and design complexity problems facing future Multiprocessor System-on-Chips(MPSoCs).Effective run-time mapping on such 3-D NoC-based MPSoCs can be quite challenging,as the arrival order and task graphs of the target applications are typically not known a priori,which can be further complicated by stringent energy requirements for NoC systems.This paper thus presents an energy-aware run-time incremental mapping algorithm(ERIM) for 3-D NoC which can minimize the energy consumption due to the data communications among processor cores,while reducing the fragmentation effect on the incoming applications to be mapped,and simultaneously satisfying the thermal constraints imposed on each incoming application.Specifically,incoming applications are mapped to cuboid tile regions for lower energy consumption of communication and the minimal routing.Fragment tiles due to system fragmentation can be gleaned for better resource utilization.Extensive experiments have been conducted to evaluate the performance of the proposed algorithm ERIM,and the results are compared against the optimal mapping algorithm(branch-and-bound) and two heuristic algorithms(TB and TL).The experiments show that ERIM outperforms TB and TL methods with significant energy saving(more than 10%),much reduced average response time,and improved system utilization.     

IMORL: Incremental Multiple-Object Recognition and Localization

This paper proposes an incremental multiple-object recognition and localization (IMORL) method. The objective of IMORL is to adaptively learn multiple interesting objects in an image. Unlike the conventional multiple-object learning algorithms, the proposed method can automatically and adaptively learn from continuous video streams over the entire learning life. This kind of incremental learning capability enables the proposed approach to accumulate experience and use such knowledge to benefit future learning and the decision making process. Furthermore, IMORL can effectively handle variations in the number of instances in each data chunk over the learning life. Another important aspect analyzed in this paper is the concept drifting issue. In multiple-object learning scenarios, it is a common phenomenon that new interesting objects may be introduced during the learning life. To handle this situation, IMORL uses an adaptive learning principle to autonomously adjust to such new information. The proposed approach is independent of the base learning models, such as decision tree, neural networks, support vector machines, and others, which provide the flexibility of using this method as a general learning methodology in multiple-object learning scenarios. In this paper, we use a neural network with a multilayer perceptron (MLP) structure as the base learning model and test the performance of this method in various video stream data sets. Simulation results show the effectiveness of this method.     

光滑公式

本文提出了光滑公式应满足的条件及三个新的光滑公式.     

增量与演化流形学习综述

流形学习的目标是发现观测数据嵌入在高维数据空间中的低维光滑流形.近年来,在线或增量地发现内在低维流形结构成为流形学习的研究热点.从增量学习和演化学习2个方面入手,对该领域已有研究进展进行综述.增量流形学习较之传统的批量流形学习方法具有动态增量的能力,而演化流形学习能够在线地发现海量动态数据的内在规律,有利于进行维数约简和数据分析.文中对主要的增量与演化流形学习算法的基本原理、特点进行了阐述,分析了各自的优点与不足,指出了该领域的开放问题,并对进一步的研究方向进行了展望.     

Online Metric Tracking and Smoothing

We consider the online smoothing problem, in which a tracker is required to maintain distance no more than Δ≥0 from a time-varying signal f while minimizing its own movement. The problem is determined by a metric space (X,d) with an associated cost function c:?→?. Given a signal f ₁,f ₂,…∈X the tracker is responsible for producing a sequence a ₁,a ₂,… of elements of X that meet the proximity constraint: d(f _i ,a _i )≤Δ. To complicate matters, the tracker is on-line—the value a _i may only depend on f ₁,…,f _i —and wishes to minimize the cost of his travels, ∑c(d(a _i ,a _i+1)). We evaluate such tracking algorithms competitively, comparing this with the cost achieved by an optimal adversary apprised of the entire signal in advance. The problem was originally proposed by Yi and Zhang (In: Proceedings of the 20th annual ACM-SIAM symposium on discrete algorithms (SODA), pp. 1098–1107. ACM Press, New York, 2009), who considered the natural circumstance where the metric spaces are taken to be ? ^k with the ? ₂ metric and the cost function is equal to 1 unless the distance is zero (thus the tracker pays a fixed cost for any nonzero motion).

We begin by studying arbitrary metric spaces with the “pay if you move” metric of Yi and Zhang (In: Proceedings of the 20th annual ACM-SIAM symposium on discrete algorithms (SODA), pp. 1098–1107. ACM Press, New York, [2009]) described above and describe a natural randomized algorithm that achieves a O(logb _Δ)-competitive ratio, where b _Δ=max _x∈X |B _Δ(x)| is the maximum number of points appearing in any ball of radius Δ. We show that this bound is tight.

We then focus on the metric space ? with natural families of monotone cost functions c(x)=x ^p for some p≥0. We consider both the expansive case (p≥1) and the contractive case (p<1), and show that the natural lazy algorithm performs well in the expansive case. In the contractive case, we introduce and analyze a novel deterministic algorithm that achieves a constant competitive ratio depending only on p. Finally, we observe that by slightly relaxing the guarantee provided by the tracker, one can obtain natural analogues of these algorithms that work in continuous metric spaces.

     

Smoothing patched grids

The purpose of this paper is to indicate how smoothing techniques can be utilized in the area of grid generation. The focus of the paper is to show how one global grid can be patched together from a number of smaller ones. The procedure usually takes place in two steps. First, one global grid is patched together from a number of smaller ones, allowing for the possibility that the derivatives along common boundaries may not be continuous. The second step is to then approximate this grid by a smooth one in such a way that the essential structure of each patch is preserved.     

Incremental critiquing

Conversational recommender systems guide users through a product space, alternatively making concrete product suggestions and eliciting the user's feedback. Critiquing is a common form of user feedback, where users provide limited feedback at the feature-level by constraining a feature's value-space. For example, a user may request a cheaper product, thus critiquing the price feature. Usually, when critiquing is used in conversational recommender systems, there is little or no attempt to monitor successive critiques within a given recommendation session. In our experience this can lead to inefficiencies on the part of the recommender system, and confusion on the part of the user. In this paper we describe an approach to critiquing that attempts to consider a user's critiquing history, as well as their current critique, when making new recommendations. We provide experimental evidence to show that this has the potential to significantly improve recommendation efficiency.     

Smoothing sample extremes: The mixed model approach

Nonparametric regression for sample extremes can be performed using a variety of techniques. The penalized spline approach for the Poisson point process model is considered. The generalized linear mixed model representation for the spline model, with its Bayesian approach to inference, turns out to be a very flexible framework. Monte Carlo Markov chain algorithms are employed for exploration of the posterior distribution. The overall performance of the method is tested on simulated data. Two real data applications are also discussed for modeling trend of intensity of earthquakes in Italy and for assessing seasonality and short term trend of summer extreme temperatures in Milan, Italy.     

Incremental Dynamics

A new system of dynamic logic is introduced and motivated, witha novel approach to variable binding for incremental interpretation. Thesystem is shown to be equivalent to first order logic and complete.The new logic combines the dynamic binding idea from DynamicPredicate Logic with De Bruijn style variable free indexing. Quantifiersbind the next available variable register; the indexing mechanismguarantees that active registers are never overwritten by newquantifiers actions. Apart from its interest in its own right, theresulting system has certain advantages over Dynamic Predicate Logic orDiscourse Representation Theory. It comes with a more well behaved(i.e., transitive) consequence relation, it gives a more explicitaccount of how anaphoric context grows as text gets processed, and ityields new insight into the dynamics of anaphoric linking in reasoning.Incremental dynamics also points to a new way of handling contextdynamically in Montague grammar.     

类别增量学习研究进展和性能评价

机器学习技术成功地应用于计算机视觉、自然语言处理和语音识别等众多领域. 然而, 现有的大多数机器学习模型在部署后类别和参数是固定的, 只能泛化到训练集中出现的类别, 无法增量式地学习新类别. 在实际应用中, 新的类别或任务会源源不断地出现, 这要求模型能够像人类一样在较好地保持已有类别知识的基础上持续地学习新类别知识. 近年来新兴的类别增量学习研究方向, 旨在使得模型能够在开放、动态的环境中持续学习新类别的同时保持对旧类别的判别能力(防止“灾难性遗忘”). 本文对类别增量学习(Class-incremental learning, CIL)方法进行了详细综述. 根据克服遗忘的技术思路, 将现有方法分为基于参数正则化、基于知识蒸馏、基于数据回放、基于特征回放和基于网络结构的五类方法, 对每类方法的优缺点进行了总结. 此外, 本文在常用数据集上对代表性方法进行了实验评估, 并通过实验结果对现有算法的性能进行了比较分析. 最后, 对类别增量学习的研究趋势进行展望.     

IncPR：一种基于增量计算的并行PageRank算法

广泛的互联网的商业应用使PageRank算法有重要地位.网络规模不断地增大,同时网络变化带来的时效性要求,也使PageRank计算对计算资源的要求不断地提高.为降低该问题对计算资源的消耗水平,降低计算成本,一种基于增量计算思想的PageRank算法：IncPR被提出.IncPR通过重用已有的结果,增量地获得数据变化后的结果.该算法在并行计算环境中,能够有效地降低计算量,缩短计算时间.理论分析表明,该算法计算结果的误差范围与蒙特卡罗PageRank算法相当,其时间复杂度优于其他已有的相关算法,且不引入额外的存储开销.在分布式集群Hama上进行的实验验证了理论分析的结果,IncPR在得到与蒙特卡罗PageRank算法同等(甚至更高)结果精度的情况下,显著地降低了计算量.     

AgentCubes: Incremental 3D end-user development

3D game development can be an enticing way to attract K-12 students to computer science, but designing and programming 3D games is far from trivial. Students need to achieve a certain level of 3D fluency in modeling, animation, and programming to be able to create compelling 3D content. The combination of innovative end-user development tools and standards-based curriculum that promotes IT fluency by shifting the pedagogical focus from programming to design, can address motivational aspects without sacrificing principled educational goals. The AgentCubes 3D game-authoring environment raises the ceiling of end-user development without raising the threshold. Our formal user study shows that with Incremental 3D, the gradual approach to transition from 2D to 3D authoring, middle school students can build sophisticated 3D games including 3D models, animations, and programming.     

Experiments with Incremental Concept Formation: UNIMEM

Learning by observation involves automatic creation of categories that summarize experience. In this paper we present UNIMEM, an artificial intelligence system that learns by observation. UNIMEM is a robust program that can be run on many domains with real-world problem characteristics such as uncertainty, incompleteness, and large numbers of examples. We give an overview of the program that illustrates several key elements, including the automatic creation of non-disjoint concept hierarchies that are evaluated over time. We then describe several experiments that we have carried out with UNIMEM, including tests on different domains (universities, Congressional voting records, and terrorist events) and an examination of the effect of varying UNIMEM's parameters on the resulting concept hierarchies. Finally we discuss future directions for our work with the program.     

分形变换与图象噪声平滑

利用分形变换进行图象噪声平滑，是他形理论的又一应用领域。本文对这一问题进行了理论和实验验证，从而指出了分形变换应用于图象噪声平滑的潜力所在。     

On Variational Curve Smoothing and Reconstruction

In this paper we discuss and experimentally compare variational methods for curve denoising, curve smoothing and curve reconstruction problems. The methods are based on defining suitable cost functionals to be minimized, the cost being the combination of a fidelity term measuring the “distance” of a curve from the data and a smoothness term measuring the curve’s L ₁-norm or length.     

Smoothing policies and safe policy gradients

Machine Learning - Policy gradient (PG) algorithms are among the best candidates for the much-anticipated applications of reinforcement learning to real-world control tasks, such as robotics....     

机器人增量学习研究综述

机器人的应用场景正在不断更新换代,数据量也在日益增长.传统的机器学习方法难以适应动态的环境,而增量学习技术能够模拟人类的学习过程,使机器人能利用旧知识来加快新任务的学习,在不遗忘旧技能的前提下学习新的技能.目前对于机器人增量学习的相关研究仍然较少,对此,主要介绍机器人增量学习研究进展.首先,对增量学习进行简介;其次,从参数和模型的角度出发,将当前机器人增量学习主流方法分为变参数方法、变模型方法、混合方法3类,分别对每一类进行论述,并给出相应的增量学习技术在机器人领域中的应用实例;然后,对机器人增量学习中常用的数据集和评价指标进行介绍;最后,对增量学习未来的发展趋势进行展望.     

基于状态的增量式类测试策略SITS

基于状态的测试是面向对象软件测试的一个重要方面。本文在FREE状态测试模型的基础上,提出了一种基于状态的增量式类测试策略。首先提出了一种全面测试基类的显式、隐式行为的测试策略;然后在进行面向对象软件测试时,对于子类的测试,加入增量式测试的思想,以简化测试实例的数量及测试执行时间;最后对SITS策略的测试实用性进进行了概括性总结。