Constrained pose refinement of parametric objects

Pose refinement is an essential task for computer vision systems that require the calibration and verification of model and camera parameters. Typical domains include the real-time tracking of objects and verification in model-based recognition systems. A technique is presented for recovering model and camera parameters of 3D objects from a single two-dimensional image. This basic problem is further complicated by the incorporation of simple bounds on the model and camera parameters and linear constraints restricting some subset of object parameters to a specific relationship. It is demonstrated in this paper that this constrained pose refinement formulation is no more difficult than the original problem based on numerical analysis techniques, including active set methods and lagrange multiplier analysis. A number of bounded and linearly constrained parametric models are tested and convergence to proper values occurs from a wide range of initial error, utilizing minimal matching information (relative to the number of parameters and components). The ability to recover model parameters in a constrained search space will thus simplify associated object recognition problems.     

Robot dynamic calibration: Optimal excitation trajectories and experimental parameter estimation

Advanced robot control schemes require an accurate knowledge of the dynamic parameters of the manipulator. This article examines various issues related to robot dynamic calibration, from generation of optimal excitation trajectories to data acquisition and filtering, and experimental inertial and friction parameter estimation. In particular, a new method is developed for the determination of optimal joint trajectories for the calibration experiment, which is based on evolutionary optimization techniques. A genetic algorithm is used to determine excitation trajectories that minimize either the condition number of the regression matrix or the logarithmic determinant of the Fisher information matrix. All the calibration steps have been carried out on a SCARA two‐link planar manipulator, and the experimental results are discussed. © 2001 John Wiley & Sons, Inc.     

基于互信息的统计语言模型平滑技术

数据平滑主要是用来解决统计语言模型在实际应用中数据稀疏问题。现有平滑技术虽然已有效地对数据稀疏问题进行了处理,但对已出现事件频率分布的合理性并没有作出有效的分析。本文则针对二元模型,提出了一种基于互信息的平滑技术,其基本思想是根据模型中每个二元对的互信息的高低对其概率进行折扣或补偿,并用极小化困惑度原则体现了模型的合理性。实验结果表明该技术优于目前常用的Katz 平滑技术。     

GPS coordinates estimation and camera calibration from solar shadows

In this paper, we discuss the issue of camera parameter estimation (intrinsic and extrinsic parameters), along with estimation of the geo-location of the camera by using only the shadow trajectories. By observing stationary objects over a period of time, it is shown that only six points on the trajectories formed by tracking the shadows of the objects are sufficient to estimate the horizon line of the ground plane. This line is used along with the extracted vertical vanishing point to calibrate the stationary camera. The method requires as few as two shadow casting objects in the scene and a set of six or more points on the shadow trajectories of these objects. Once camera intrinsic parameters are recovered, we present a novel application where one can accurately determine the geo-location of the camera up to a longitude ambiguity using only three points from these shadow trajectories without using any GPS or other special instruments. We consider possible cases where this ambiguity can also be removed if additional information is available. Our method does not require any knowledge of the date or the time when the images are taken, and recovers the date of acquisition directly from the images. We demonstrate the accuracy of our technique for both steps of calibration and geo-temporal localization using synthetic and real data.     

An information theoretic approach to camera control for crowded scenes

Navigation and monitoring of large and crowded virtual environments is a challenging task and requires intuitive camera control techniques to assist users. In this paper, we present a novel automatic camera control technique providing a scene analysis framework based on information theory. The developed framework contains a probabilistic model of the scene to build entropy and expectancy maps. These maps are utilized to find interest points which represent either characteristic behaviors of the crowd or novel events occurring in the scene. After an interest point is chosen, the camera is updated accordingly to display this point. We tested our model in a crowd simulation environment and it performed successfully. Our method can be integrated into existent camera control modules in computer games, crowd simulations and movie pre-visualization applications.     

Face model fitting with learned displacement experts and multi-band images

Analyzing human faces is a traditional topic in computer vision research. For this task, model based approaches have been proven adequate to extract high-level information in many applications. However, they require a robust estimation of model parameters to work reliably. To tackle this challenge, we train displacement experts that serve as an update function on initial model parameter configurations. Unfortunately, building displacement experts that work robustly even in unconstrained environments is a non-trivial task. Therefore, we rely on a priori information about the structure of human faces by integrating an image representation that reflects the location of several facial components, so called “multi-band images”. By combining multi-band images and learned displacement experts, we propose a novel face model fitting approach. An evaluation on the “Labeled Faces In The Wild” database demonstrates, that this approach provides robust fitting results even in unconstrained environments.     

A new approach to inversion-based feedforward control design for nonlinear systems

The finite-time transition between stationary setpoints of nonlinear SISO systems is considered as a scenario for the presentation of a new design approach for inversion-based feedforward control. Design techniques which are based on a stable system inversion result in input trajectories with pre- and/or post-actuation intervals. The presented approach treats the considered transition task as a two-point boundary value problem (BVP) and yields causal feedforward trajectories, which are constant outside the transition interval. The main idea of this approach is to provide free parameters in the desired output trajectory to solve the BVP of the internal dynamics. Thereby, a standard MATLAB function can be used for the numerical solution of the BVP. Feedforward control design techniques are illustrated by simulation results for a simple example.     

Open-loop stochastic control of pharmacokinetic systems: a new method for design of dosing regimens.

On the basis stochastic control techniques, an algorithm for the design of dosing regimens is developed. The essence of the method relies on a constrained model for the population and on a first-order approximation in the evaluation of the performance cost function. Since it does not require detailed information on the probability distribution of the model parameters, the method can be used in a wide range of practical applications. The application of the design technique is illustrated in two typical situations: lacking a probability model for the pharmacokinetic parameters (etomidate) and having a probability model for the pharmacokinetic parameters (theophylline).     

Examining the costs of multiple trajectory pointing techniques

Pointing techniques that offer users multiple trajectories to a target have the potential to reduce pointing time by allowing a shorter than normal movement distance. However, such techniques potentially introduce additional elements into the pointing task involving identification of the alternative trajectories, assessment of their relative performance, and selection of the one to use. These additional tasks may reduce or negate the benefits of offering shorter paths. To better understand these issues, we developed a methodology for controlling the relative benefits of alternative target trajectories, and used it to evaluate three interfaces—a ‘pointer wrapping’ technique that allows the cursor to traverse from one screen edge to the opposing edge, and a system allowing users to choose between multiple cursors in two configurations (‘Ninja cursors’). We found that performance with these techniques was significantly worse than that predicted by Fitts' law for a single cursor, suggesting that the additional elements in their use are significant. Detailed analysis of behaviour during acquisitions showed that much of this cost is accrued in the mental preparation that precedes motor action, and in additional volatility in the pointing movements. We discuss how the method and findings may assist those developing enhanced pointing techniques.     

Online monitoring by dynamically refining imprecise models

Model-based monitoring determines faults in a supervised system by comparing the available system's measurements with a priori information represented by the system's mathematical model. Especially in technical environments, a monitoring system must be able to reason with incomplete knowledge about the supervised system, to process noisy and erroneous observations and to react within a limited time. We present MOSES, a model-based monitoring system which is based on imprecise models where the structure is known and the parameters may be imprecisely specified by numerical intervals. As a consequence, only bounds on the trajectories can be derived with imprecise models. These bounds are computed using traditional numerical integration techniques starting from individual points on the external surface of the model's uncertainty space. When new measurements from the supervised system become available, MOSES checks the consistency of this new information with the model's prediction and refutes inconsistent parts from the uncertainty space of the model. A fault in the supervised system is detected when the complete model's uncertainty space has been refuted. MOSES bridges and extends methodologies from the FDI and DX communities by refining the model's uncertainty space conservatively through refutation, by applying standard numerical techniques for deriving the trajectories of imprecise models and by exploiting the measurements as soon as possible for online monitoring. The performance of MOSES is evaluated based on examples and by online monitoring a complex heating system.     

Human gesture recognition system for TV viewing using time-of-flight camera

We developed a new device-free user interface for TV viewing that uses a human gesture recognition technique. Although many motion recognition technologies have been reported, no man–machine interface that recognizes a large enough variety of gestures has been developed. The difficulty was the lack of spatial information that could be acquired from normal video sequences. We overcame the difficulty by using a time-of-flight camera and novel action recognition techniques. The main functions of this system are gesture recognition and posture measurement. The former is performed using the bag-of-features approach, which uses key-point trajectories as features. The use of 4-D spatiotemporal trajectory features is the main technical contribution of the proposed system. The latter is obtained through face detection and object tracking technology. The interface is useful because it does not require any contact-type devices. Several experiments proved the effectiveness of our proposed method and the usefulness of the system.     

A new evolutionary algorithm combining simulated annealing and genetic programming for relevance feedback in fuzzy information retrieval systems

 Relevance feedback techniques have demonstrated to be a powerful means to improve the results obtained when a user submits a query to an information retrieval system as the world wide web search engines. These kinds of techniques modify the user original query taking into account the relevance judgements provided by him on the retrieved documents, making it more similar to those he judged as relevant. This way, the new generated query permits to get new relevant documents thus improving the retrieval process by increasing recall. However, although powerful relevance feedback techniques have been developed for the vector space information retrieval model and some of them have been translated to the classical Boolean model, there is a lack of these tools in more advanced and powerful information retrieval models such as the fuzzy one. In this contribution we introduce a relevance feedback process for extended Boolean (fuzzy) information retrieval systems based on a hybrid evolutionary algorithm combining simulated annealing and genetic programming components. The performance of the proposed technique will be compared with the only previous existing approach to perform this task, Kraft et al.'s method, showing how our proposal outperforms the latter in terms of accuracy and sometimes also in time consumption. Moreover, it will be showed how the adaptation of the retrieval threshold by the relevance feedback mechanism allows the system effectiveness to be increased.     

Towards A New Authoring Methodology for Large-Scale Hypermedia Applications

As the amount of information technology increases, managing information resources, so that the correct people can find the information easily, becomes a critical issue. Hypermedia systems are considered one solution to this problem as they provide a means for representing higher level relationships between the underlying information. However, the amount of information available electronically is increasing at an accelerated rate. Using standard hypermedia authoring techniques, the effort required managing and maintaining large-scale hypermedia systems is enormous. Hypermedia authoring in the large requires new methodologies if it is going to be feasible.This paper presents a new model for building and structuring large-scale hypermedia applications. It describes a case study that explored the delivery of hypermedia information in an industrial environment on a small scale. Models and techniques developed for that case study were then refined and augmented so they could support the construction of large-scale hypermedia systems. In order to support such endeavors a new linking model is presented that allows the author to explicitly represent abstract concepts contained within the underlying information and interconnect them in some meaningful manner. A prototype system applying this linking technique is presented.     

Variable selection procedures and efficient suboptimal mask search algorithms in fuzzy inductive reasoning

This paper describes two new suboptimal mask search algorithms for Fuzzy inductive reasoning (FIR), a technique for modelling dynamic systems from observations of their input/output behaviour. Inductive modelling is by its very nature an optimisation problem. Modelling large-scale systems in this fashion involves solving a high-dimensional optimisation problem, a task that invariably carries a high computational cost. Suboptimal search algorithms are therefore important. One of the two proposed algorithms is a new variant of a directed hill-climbing method. The other algorithm is a statistical technique based on spectral coherence functions. The utility of the two techniques is demonstrated by means of an industrial example. A garbage incinerator process is inductively modelled from observations of 20 variable trajectories. Both suboptimal search algorithms lead to similarly good models. Each of the algorithms carries a computational cost that is in the order of a few percent of the cost of solving the complete optimisation problem. Both algorithms can also be used to filter out variables of lesser importance, i.e. they can be used as variable selection tools.     

Reduction of stable differential–algebraic equation systems via projections and system identification

Most large-scale process models derived from first principles are represented by nonlinear differential–algebraic equation (DAE) systems. Since such models are often computationally too expensive for real-time control, techniques for model reduction of these systems need to be investigated. However, models of DAE type have received little attention in the literature on nonlinear model reduction. In order to address this, a new technique for reducing nonlinear DAE systems is presented in this work. This method reduces the order of the differential equations as well as the number and complexity of the algebraic equations. Additionally, the algebraic equations of the resulting system can be replaced by an explicit expression for the algebraic variables such as a feedforward neural network. This last property is important insofar as the reduced model does not require a DAE solver for its solution but system trajectories can instead be computed with regular ODE solvers. This technique is illustrated with a case study where responses of several different reduced-order models of a distillation column with 32 differential equations and 32 algebraic equations are compared.     

Generating process model collections

Business process management plays an important role in the management of organizations. More and more organizations describe their operations as business processes. It is common for organizations to have collections of thousands of business processes, but for reasons of confidentiality these collections are often not, or only partially, available to researchers. On the other hand, research on techniques for managing process model collections, such as techniques for process retrieval, requires large collections for evaluation purposes. Therefore, this paper proposes a technique to generate such collections of process models, based on the properties of real-world collections. Where existing techniques focus on the structure of the process models, the technique proposed in this paper also generates task labels that consists of words from real-life task labels and considers semantic information of node and edge types. We evaluate our technique by applying it to generate two synthetic collections of process models of over 60,000 and over 2,000 models, respectively. We show that the generated synthetic collections have similar properties to the original collections. To the best of our knowledge, this is the first technique that can generate synthetic BPMN models, thus enabling experimentation with process collections that have laboratory-set quantitative parameters and qualitative properties that are based on real-world process model collections.     

An agent-based paradigm for the reconstruction of conical perspectives

Obtaining three-dimensional models from digital images is one of the most important challenges in the computer vision. This task requires the advances in two major investigation fields: on the one hand the vectorization of the digital image to turn bitmaps into a set of geometrical entities (graph), on the other hand the reconstruction of the three-dimensional model from these two-dimensional entities. This work is focused on the second task, that is, the reconstruction of a three-dimensional model from a graph.We have focused on the reconstruction of graphs with parallel and oblique perspectives with two vanishing points, and we propose an algorithm in order to identify this typology in contrast to the proposals kwon so far, in which it is assumed that the graph always represents an oblique perspective with three vanishing points from which greater information of the three-dimensional model can be extracted.On the other hand, the current reconstruction algorithms require a previous knowledge of the faces defined in the model which must be obtained from the starting graph, task that is not currently solved completely, generates many erroneous results, requires interacting with the user and involves a high computing cost.The reconstruction algorithms proposed in this work do not require the knowledge of the faces of the model, it is completely automatic and we have obtained a success rate of 100% in the graphs with normalon and quasi-normalon typology tested.Regarding to practical advantages and research contribution in expert and intelligent systems, we propose, for the first time, an architecture based on agents for the reconstruction which allows, on the one hand, the simultaneous reconstruction of graphs considerably reducing the computing cost and, on the other hand, adding, with less effort, new reconstruction techniques to deal with graphs with other typologies.     

Optimal motion coordination of two robots—a polynomial parameterization approach to trajectory resolution

A novel approach is presented for the optimal coordination of two robots performing continuous-path manufacturing tasks. Two types of operations are considered: contact operations, such as deburring, which require the tool to maintain contact with the workpiece during the operation execution, or non-contact operations, such as arc-welding, where the tool moves relative to the workpiece without contact. In both cases, the required tool-path is specified with respect to the workpiece. A task is performed by mounting the tool on a robot, while a second robot grips the workpiece; the two robots are then coordinated to move simultaneously relative to one another so that the tool follows its prescribed trajectory at a constant speed relative to the workpiece, and provides sufficient contact force in the case of contact operations. The original tool-trajectory is thus resolved into a pair of conjugate trajectories, specified in task space relative to an inertial frame, describing the motion of the two robots. In cases where the two robots form a kinematically redundant system, the trajectory resolution process can yield an infinity of conjugate-trajectory pairs corresponding to a given original tool-trajectory, of which one pair may be chosen based on a specific choice of cost function. This article presents a technique whereby the robot's conjugate trajectories are parameterized using polynomial functions. A method is then developed for optimizing the robot's motion by selecting the optimal pair of conjugate trajectories. The proposed optimal trajectory-resolution technique is further enhanced by coupling it to a procedure for selecting the optimal relative placement of the two robots, resulting in the best achievable solution. This method of coordinating two robots yields lower cycle-time values than the corresponding single-robot operation, as well as reduces the need for complex jigs and fixtures. Numerical simulations are presented to illustrate the proposed technique. © 1994 John Wiley & Sons, Inc.     

基于概念图的相关反馈技术研究

相关反馈技术是信息检索技术研究的热点。目前常用的反馈技术依然是基于关键词匹配的方式,基于语义的方式缺少概念之间关系的描述。文章提出了一种基于概念图的相关反馈技术,采用概念图的知识表示方式描述概念之间关系,从语义的层次上进行相似度判断,扩展查询式。实验表明该方法可以满足用户的需求,提高检索的效率。