Complementary models and smoothing

The authors apply the concept of complementary models introduced by Weinert and Desai (1983) to derive forwards and backwards Markovian models for the smoothing error process. By exploring the structure of the complementary models they show that, under certain restrictions, only two simple structured models exist, one that runs forwards in time and another than runs backwards in time. The forwards complementary model leads to a forwards Rauch-Tung-Striebel (RTS) smoothing formula and to a backwards Markovian model for the error, whereas the backwards model leads to a backwards RTS formula and to a forwards error model. The two models for the smoothing error can be derived one from the other by a forwards-backwards transformation that preserves the sample paths. Finally, by using a combination of the two complementary models another proof of the two-filter smoothing formula is given     

A unified approach to smoothing formulas

Based on various approaches, several different solutions to the smoothing problem have been given. The relationships between these solutions are not immediate, although they solve the same problem. Making use of a certain framework from scattering theory, we derive two families of solutions, with equations evolving forwards and backwards in time, respectively. Within these families three major previous approaches are obtained as special cases, and their relationships are clarified. The set of solutions also contains as a fourth special case a (new) backwards analog of the innovations solution. The Mayne-Fraser two-filter formula belongs to the set of backwards solutions, and within this framework certain difficulties with its interpretation can be resolved.     

Natural deduction as higher-order resolution

An interactive theorem prover, Isabelle, is under development. In lcf, each inference rule is represented by one function for forwards proof and another (a tactic) for backwards proof. In Isabelle, each inference rule is represented by a Horn clause. Resolution gives both forwards and backwards proof, supporting a large class of logics. Isabelle has been used to prove theorems in Martin-Löf's constructive type theory. Quantifiers pose several difficulties: substitution, bound variables, Skolemization. Isabelle's representation of logical syntax is the typed λ-calculus, requiring higher-order unification. It may have potential for logic programming. Depth-first subgoaling along inference rules constitutes a higher-order PROLOG.     

New square-root smoothing algorithms

This paper presents new square-root smoothing algorithms for the three best-known smoothing formulas: (1) Rauch-Tung-Striebel (RTS) formulas, (2) Desai-Weinert-Yusypchuk (DWY) formulas, called backward RTS formulas, and (3) Mayne-Fraser (MF) formulas, called two-filter formulas. The main feature of the new algorithms is that they use unitary rotations to replace all matrix inversion and backsubstitution steps common in earlier algorithms with unitary operations; this feature enables more efficient systolic array and parallel implementations and leads to algorithms with better numerical stability and conditioning properties     

Using efference copy and a forward internal model for adaptive biped walking

To behave properly in an unknown environment, animals or robots must distinguish external from self-generated stimuli on their sensors. The biologically inspired concepts of efference copy and internal model have been successfully applied to a number of robot control problems. Here we present an application of this for our dynamic walking robot RunBot. We use efference copies of the motor commands with a simple forward internal model to predict the expected self-generated acceleration during walking. The difference to the actually measured acceleration is then used to stabilize the walking on terrains with changing slopes through its upper body component controller. As a consequence, the controller drives the upper body component (UBC) to lean forwards/backwards as soon as an error occurs resulting in dynamical stable walking. We have evaluated the performance of the system on four different track configurations. Furthermore we believe that the experimental studies pursued here will sharpen our understanding of how the efference copies influence dynamic locomotion control to the benefit of modern neural control strategies in robots.     

Kneser–Ney Smoothing With a Correcting Transformation for Small Data Sets

We present a technique which improves the Kneser-Ney smoothing algorithm on small data sets for bigrams, and we develop a numerical algorithm which computes the parameters for the heuristic formula with a correction. We give motivation for the formula with correction on a simple example. Using the same example, we show the possible difficulties one may run into with the numerical algorithm. Applying the algorithm to test data we show how the new formula improves the results on cross-entropy.     

一种改进的隐马尔可夫模型在语音识别中的应用

提出了一种新的马尔可夫模型——异步隐马尔可夫模型．该模型针对噪音环境下语音识别过程中出现丢失帧的情况，通过增加新的隐藏时间标示变量Ck，估计出实际观察值对应的状态序列，实现对不规则或者不完整采样数据的建模．详细介绍了适合异步HMM的前后向算法以及用于训练的EM算法，并且对转移矩阵的计算进行了优化．最后通过实验仿真，分别使用经典HMM和异步HMM对相同的随机抽取帧的语音数据进行识别,识别结果显示在抽取帧相同情况下异步HMM比经典HMM的识别错误率低．     

Smoothing for multipoint boundary value models

We show how to solve the linear least-squares smoothing problem for multipoint boundary value models. The complementary model is derived and is used to determine the Hamiltonian equations for the smoothed state estimate and its error covariance. Stable algorithms are obtained using an invariant imbedding/multiple shooting procedure.     

采用分段RTS的CPHD平滑算法

针对多目标跟踪中的固定间隔平滑问题,将势概率假设密度（CPHD）滤波器和RTS平滑器相结合,提出了RTS的势概率假设密度滤波平滑算法。考虑到在平滑过程中存在较大的输出延迟问题,采用分段思想,提出了分段RTS的势概率假设密度滤波平滑算法。对需要平滑的估计值进行分段;采用匈牙利算法进行航迹-估计关联;对关联后的估计值逐段进行RTS平滑。实验结果表明,与CPHD滤波结果相比,分段RTS的势概率假设密度滤波平滑算法能够更加精确地估计目标状态,并且可以有效避免直接应用RTS平滑造成的实时性欠佳问题。     

The Disconnection Tableau Calculus

In this paper we give a comprehensive presentation of the disconnection tableau calculus, a proof method for formulas in classical first-order clause logic. The distinguishing property of this calculus is that it uses unification in such a manner that important proof-theoretic advantages of the classical (i.e., Smullyan-style) tableau calculus are preserved, specifically the termination and model generation characteristics for certain formula classes. Additionally, the calculus is well suited for fully automated proof search. The calculus is described in detail with soundness and completeness proofs, and a number of important calculus refinements developed over the past years are presented. Referring to the model-finding abilities of the disconnection calculus, we explain the extraction and representation of models. We also describe the integration of paramodulation-based equality handling. Finally, we give an overview of related methods.     

A Study on Smoothing for Particle-Filtered 3D Human Body Tracking

Stochastic models have become the dominant means of approaching the problem of articulated 3D human body tracking, where approximate inference is employed to tractably estimate the high-dimensional (～30D) posture space. Of these approximate inference techniques, particle filtering is the most commonly used approach. However filtering only takes into account past observations—almost no body tracking research employs smoothing to improve the filtered inference estimate, despite the fact that smoothing considers both past and future evidence and so should be more accurate. In an effort to objectively determine the worth of existing smoothing algorithms when applied to human body tracking, this paper investigates three approximate smoothed-inference techniques: particle-filtered backwards smoothing, variational approximation and Gibbs sampling. Results are quantitatively evaluated on both the HumanEva dataset as well as a scene containing occluding clutter. Surprisingly, it is found that existing smoothing techniques are unable to provide much improvement on the filtered estimate, and possible reasons as to why are explored and discussed.     

Bisimulation is two-way simulation

We give here a simple proof of the fact that on transition systems bisimulation is the equivalence relation generated by simulation via functions. The proof entirely rests on simple rules of the calculus of relations.     

基于噪点去除的人脸三维网格光顺

文章针对人脸这一特殊模型的去噪和光顺问题,结合网格平滑的常用策略,从模型组织,噪点检测与修正,网格平滑等方面给出了具体的解决方案。经实践证明结果完全符合要求。     

Representing model theory in a type-theoretical logical framework

In a broad sense, logic is the field of formal languages for knowledge and truth that have a formal semantics. It tends to be difficult to give a narrower definition because very different kinds of logics exist. One of the most fundamental contrasts is between the different methods of assigning semantics. Here two classes can be distinguished: model theoretical semantics based on a foundation of mathematics such as set theory, and proof theoretical semantics based on an inference system possibly formulated within a type theory.Logical frameworks have been developed to cope with the variety of available logics unifying the underlying ontological notions and providing a meta-theory to reason abstractly about logics. While these have been very successful, they have so far focused on either model or proof theoretical semantics. We contribute to a unified framework by showing how the type/proof theoretical Edinburgh Logical Framework (LF) can be applied to the representation of model theoretical logics.We give a comprehensive formal representation of first-order logic, covering both its proof and its model theoretical semantics as well as its soundness in LF. For the model theory, we have to represent the mathematical foundation itself in LF, and we provide two solutions for that. Firstly, we give a meta-language that is strong enough to represent the model theory while being simple enough to be treated as a fragment of untyped set theory. Secondly, we represent Zermelo-Fraenkel set theory and show how it subsumes our meta-language. Specific models are represented as LF morphisms.All representations are given in and mechanically verified by the Twelf implementation of LF. Moreover, we use the Twelf module system to treat all connectives and quantifiers independently. Thus, individual connectives are available for reuse when representing other logics, and we obtain the first version of a feature library from which logics can be pieced together.Our results and methods are not restricted to first-order logic and scale to a wide variety of logical systems, thus demonstrating the feasibility of comprehensively formalizing large scale representation theorems in a logical framework.     

Smart cards: The Authenticated Solution For E-business User

Most business processes have traditionally been paper-based and confirmed in writing. However, the advent of the Internet means that business transactions can be undertaken far more efficiently — and by that I mean electronically. In fact, one of the key advantages of electronic information is that it can be incorporated into business directly, without the unnecessary passing backwards and forwards of reams of information.     

Certificateless public key encryption: A new generic construction and two pairing-free schemes

The certificateless encryption (CLE) scheme proposed by Baek, Safavi-Naini and Susilo is computation-friendly since it does not require any pairing operation. Unfortunately, an error was later discovered in their security proof and so far the provable security of the scheme remains unknown. Recently, Fiore, Gennaro and Smart showed a generic way (referred to as the FGS transformation) to transform identity-based key agreement protocols to certificateless key encapsulation mechanisms (CL-KEMs). As a typical example, they showed that the pairing-free CL-KEM underlying Baek et al.’s CLE can be “generated” by applying their transformation to the Fiore–Gennaro (FG) identity-based key agreement (IB-KA) protocol.In this paper, we show that directly applying the Fiore–Gennaro–Smart (FGS) transformation to the original FG IB-KA protocol in fact results in an insecure CL-KEM scheme against strong adversaries, we also give a way to fix the problem without adding any computational cost. The reason behind our attack is that the FGS transformation requires the underlying IB-KA protocol to be secure in a model that is stronger than the conventional security models where existing IB-KA protocols are proved secure, and the FG IB-KA protocol is in fact insecure in the new model. This motivates us to construct a new generic transformation from IB-KA protocols to CLE schemes. In the paper we present such a transformation which only requires the underlying IB-KA protocol to be secure in a security model that is weaker than the existing security models for IB-KA protocols. We illustrate our transformation by generating a new pairing-free CLE scheme that is obtained by directly applying our transformation to the original FG IB-KA protocol.     

Scaling Smoothed Language Models

In Continuous Speech Recognition (CSR) systems a Language Model (LM) is required to represent the syntactic constraints of the language. Then a smoothing technique needs to be applied to avoid null LM probabilities. Each smoothing technique leads to a different LM probability distribution. Test set perplexity is usually used to evaluate smoothing techniques but the relationship with acoustic models is not taken into account. In fact, it is well-known that to obtain optimum CSR performances a scaling exponential parameter must be applied over LMs in the Bayes’ rule. This scaling factor implies a new redistribution of smoothed LM probabilities. The shape of the final probability distribution is due to both the smoothing technique used when designing the language model and the scaling factor required to get the optimum system performance when integrating the LM into the CSR system. The main object of this work is to study the relationship between the two factors, which result in dependent effects. Experimental evaluation is carried out over two Spanish speech application tasks. Classical smoothing techniques representing very different degrees of smoothing are compared. A new proposal, Delimited discounting, is also considered. The results of the experiments showed a strong dependence between the amount of smoothing given by the smoothing technique and the way that the LM probabilities need to be scaled to get the best system performance, which is perplexity independent in many cases. This relationship is not independent of the task and available training data.     

Avoiding duplicate proofs with the foothold refinement

Wos has identified the problem of recomputing redundant information in the general setting of automated reasoning. We consider this problem in the setting of logic programming where we are given a formula and a goal and asked to find the instances of the goal that follow from the formula. We can use a proof procedure to find the result. The procedure exhibits redundancy when it finds two results such that one either duplicates or is more specific than the other. We introduce the foothold format, a refinement of linear resolution that admits fewer duplicate proofs than Loveland's popular MESON format. The duplicates arise when reasoning by cases. It leads to proof procedures that compute fewer duplicate substitutions. In some of our examples all duplication is eliminated. The foothold refinement depends on a simple condition that can be checked quickly, and can detect redundancy before the proof is completely generated. This is important from a practical point of view, since the earlier redundancy is detected, the more unnecessary work can be avoided. For some examples the speedup is exponential. We show that the elimination of redundancy also applies to SLI resolution, a procedure for processing disjunctive logic programs.     

基于分段RTS平滑的凸组合航迹融合算法

针对凸组合航迹融合算法在过程噪声不为零的情况下性能下降的问题,引入了RTS平滑算法来提高融合性能。由于传统的RTS平滑算法是得到全部滤波结果之后才执行逆向平滑过程,造成输出延迟,为此,提出了分段RTS平滑算法,一方面可以提高航迹融合性能,另一方面能够保证融合过程中的实时性。在融合过程中,针对局部节点有无额外计算能力的不同情况,结合实施平滑步骤的时机,提出了基于分段RTS平滑的先平滑再融合和先融合再平滑两种改进的凸组合航迹融合方法。这两种方法在不同过程噪声水平下,性能表现都超过凸组合融合算法和最优融合算法。仿真结果表明了新算法的有效性和优越性。     

六足移动式微型仿生机器人的研究

本文描述了一种微型六足仿生机器人的结构与控制，分析了这种微型六足仿生机器人 的移动原理. 该机器人基于仿生学原理，结构独特、简单、新颖，能方便地实现前进和后退 ，其样机外形尺寸为：长30mm，宽40mm，高20mm，重6.3g．并对该样机进行了实验，实验 结果表明该机器人具有较好的机动性．