Multi-Mode Tensor Representation of Motion Data

In this paper, we investigate how a multilinear model can be used to represent human motion data. Based on technical modes (referring to degrees of freedom and number of frames) and natural modes that typically appear in the context of a motion capture session (referring to actor, style, and repetition), the motion data is encoded in form of a high-order tensor. This tensor is then reduced by using N-mode singular value decomposition. Our experiments show that the reduced model approximates the original motion better then previously introduced PCA-based approaches. Furthermore, we discuss how the tensor representation may be used as a valuable tool for the synthesis of new motions.

Simultaneous Learning of Nonlinear Manifold and Dynamical Models for High-dimensional Time Series

The goal of this work is to learn a parsimonious and informative representation for high-dimensional time series. Conceptually, this comprises two distinct yet tightly coupled tasks: learning a low-dimensional manifold and modeling the dynamical process. These two tasks have a complementary relationship as the temporal constraints provide valuable neighborhood information for dimensionality reduction and conversely, the low-dimensional space allows dynamics to be learnt efficiently. Solving these two tasks simultaneously allows important information to be exchanged mutually. If nonlinear models are required to capture the rich complexity of time series, then the learning problem becomes harder as the nonlinearities in both tasks are coupled. The proposed solution approximates the nonlinear manifold and dynamics using piecewise linear models. The interactions among the linear models are captured in a graphical model. By exploiting the model structure, efficient inference and learning algorithms are obtained without oversimplifying the model of the underlying dynamical process. Evaluation of the proposed framework with competing approaches is conducted in three sets of experiments: dimensionality reduction and reconstruction using synthetic time series, video synthesis using a dynamic texture database, and human motion synthesis, classification and tracking on a benchmark data set. In all experiments, the proposed approach provides superior performance.

Simultaneous Learning of Non-Linear Manifold and Dynamical Models for High-Dimensional Time Series

The goal of this work is to learn a parsimonious and informative representation for high-dimensional time series. Conceptually, this comprises two distinct yet tightly coupled tasks: learning a low-dimensional manifold and modeling the dynamical process. These two tasks have a complementary relationship as the temporal constraints provide valuable neighborhood information for dimensionality reduction and conversely, the low-dimensional space allows dynamics to be learnt efficiently. Solving these two tasks simultaneously allows important information to be exchanged mutually. If nonlinear models are required to capture the rich complexity of time series, then the learning problem becomes harder as the nonlinearities in both tasks are coupled. The proposed solution approximates the nonlinear manifold and dynamics using piecewise linear models. The interactions among the linear models are captured in a graphical model. The model structure setup and parameter learning are done using a variational Bayesian approach, which enables automatic Bayesian model structure selection, hence solving the problem of over-fitting. By exploiting the model structure, efficient inference and learning algorithms are obtained without oversimplifying the model of the underlying dynamical process. Evaluation of the proposed framework with competing approaches is conducted in three sets of experiments: dimensionality reduction and reconstruction using synthetic time series, video synthesis using a dynamic texture database, and human motion synthesis, classification and tracking on a benchmark data set. In all experiments, the proposed approach provides superior performance.

Confirmação do modelo da estrutura 3D do vale inferior do Tejo a partir de dados de ruído sísmico ambiente

Ao longo de sua história a região do Vale Inferior do Tejo VIT foi abalada por vários sismos consideravelmente destrutivas, tendo alguns deles produzido significativas deformações nas estruturas marítimas localizadas no litoral a sudoeste do território Português; outros, moderados, foram produzidos por fontes locais, como os de 1344, 1531 e 1909. Nos últimos anos, devido à melhoria dos modelos de estrutura 3D e o desenvolvimento dos métodos numéricos, foram elaborados vários estudos de síntese de movimento forte do solo para a região do Baixo Tejo utilizando o método de diferenças finitas. Para confirmar o modelo de velocidades desta bacia usámos medidas de ruído sísmico, aplicámos um método baseado na razão espectral H/V, e, a partir destas curvas, por inversão, obtivemos um modelo de velocidades para a região estudada. Os resultados revelam uma boa concordância entre o modelo obtido e os dados geofísicos e geológicos recolhidos na mesma área._ ABSTRACT: Along his history the Lower Tagus Valley (LTV) area was shaken by several earthquakes. The largest reported had their origin in the southwestern part of Iberia. Other moderate earthquakes were produced by local sources such as the 1344, 1531 and the 1909. ln the last years, due to 3D structural model improvement and development in numerical methods, several studies have successful obtained strong-ground motion synthesis for the LVT region using finite difference method. To confirm the velocity model of the LTV sedimentary basin obtained by geophysical and geological data, we use broad-band microtremor measurements and application of the horizontal to vertical (H/V) spectral ratio method. We have obtained a velocity model for the studied region by inversion of the H/V curve. The results have good agreement geological and geophysical data.

Continuous passive motion applied to whole joints stimulates chondrocyte blosynthesis of PRG4

Continuous passive motion (CPM) is currently a part of patient rehabilitation regimens after a variety of orthopedic surgical procedures. While CPM can enhance the joint healing process, the direct effects of CPM on cartilage metabolism remain unknown. Recent in vivo and in vitro observations suggest that mechanical stimuli can regulate articular cartilage metabolism of proteoglycan 4 (PRG4), a putative lubricating and chondroprotective molecule found in synovial fluid and at the articular cartilage surface. ----- ----- Objectives: (1) Determine the topographical variation in intrinsic cartilage PRG4 secretion. (2) Apply a CPM device to whole joints in bioreactors and assess effects of CPM on PRG4 biosynthesis.----- ----- Methods: A bioreactor was developed to apply CPM to bovine stifle joints in vitro. Effects of 24 h of CPM on PRG4 biosynthesis were determined.----- ----- Results: PRG4 secretion rate varied markedly over the joint surface. Rehabilitative joint motion applied in the form of CPM regulated PRG4 biosynthesis, in a manner dependent on the duty cycle of cartilage sliding against opposing tissues. Specifically, in certain regions of the femoral condyle that were continuously or intermittently sliding against meniscus and tibial cartilage during CPM, chondrocyte PRG4 synthesis was higher with CPM than without.----- ----- Conclusions: Rehabilitative joint motion, applied in the form of CPM, stimulates chondrocyte PRG4 metabolism. The stimulation of PRG4 synthesis is one mechanism by which CPM may benefit cartilage and joint health in post-operative rehabilitation. (C) 2006 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved.

Deblurring in a noncoherent optical processing system: pupil function synthesis and experimental implementation

The bipolar point spread function (PSF) corresponding to the Wiener filter tor correcting linear-motion-blurred pictures is implemented in a noncoherent optical processor. The following two approaches are taken for this implementation: (1) the PSF is modulated and biased so that the resulting function is non-negative and (2) the PSF is split into its positive and sign-reversed negative parts, and these two parts are dealt with separately. The phase problem associated with arriving at the pupil function from these modified PSFs is solved using both analytical and combined analytical-iterative techniques available in the literature. The designed pupil functions are experimentally implemented, and deblurring in a noncoherent processor is demonstrated. The postprocessing required (i.e., demodulation in the first approach to modulating the PSF and intensity subtraction in the second approach) are carried out either in a coherent processor or with the help of a PC-based vision system. The deblurred outputs are presented.

Synthesis of path generating flexible-link mechanisms

Flexible-link mechanisms are those linkage mechanisms (or structures) which are capable of motion by virtue of elastic deformation of one or more;links. In such mechanisms a single flexible link; can replace several rigid links and joints resulting in fewer links, fewer pin joints, reduced overall weight and reduced mechanical error. In spite of such clear advantages, contributions toward flexible-link mechanisms remain very scarce. The area of flexible-link mechanisms offers much scope for further exploration. This paper attempts to show the potential of flexible-link mechanisms in accomplishing a kinematic task like path generation. Synthesis of a four-bar mechanism with a flexible rocker for circular and straight line path generation is carried out. Displacement analysis of the structure is carried out using finite element method (FEM) and synthesis is formulated and solved as an optimization problem. Several numerical examples are presented for illustration. Based on the results obtained with these examples, the flexible-link mechanism considered shows good promise for-path generation.

KINEMATIC SYNTHESIS AND ANALYSIS OF THE RACK AND PINION MULTIPURPOSE MECHANISM

The general procedure for synthesizing the rack and pinion mechanism up to seven precision conditions is developed. To illustrate the method, the mechanism has been synthesized in closed form for three precision conditions of path generation, two positions of function generation, and a velocity condition at one of the precision points. This mechanism has a number of advantages over conventional four bar mechanisms. First, since the rack is always tangent to the pinion, the transmission angle is always 90 deg minus the pressure angle of the rack. Second, with both translation and rotation of the rock occurring, multiple outputs are available. Other advantages include the generation of monotonic functions for a wide variety of motion and nonmonotonic functions for a full range of motion as well as nonlinear amplified motions. In this work the mechanism is made to satisfy a number of amplified motions. In this work the mechanism is made to satisfy a number of practical design requirements such as completely rotatable input crank and others. By including the velocity specification, the designer has considerably more control of the output motion. The method of solution developed in this work uses the complex number method of mechanism synthesis. A numerical example is included

Kinematic Synthesis and Analysis of the Rack and Pinion Multi-Purpose Mechanism

The general procedure for synthesizing the rack and pinion mechanism up to seven precision conditions is developed. To illustrate the method, the mechanism has been synthesized in closed form for three precision conditions of path generation, two positions of function generation, and a velocity condition at one of the precision points. This mechanism has a number of advantages over conventional four bar mechanisms. First, since the rack is always tangent to the pinion, the transmission angle is always 90 deg minus the pressure angle of the rack. Second, with both translation and rotation of the rack occurring, multiple outputs are available. Other advantages include the generation of monotonic functions for a wide variety of motion and nonmonotonic functions for a full range of motion as well as nonlinear amplified motions. In this work the mechanism is made to satisfy a number of practical design requirements such as completely rotatable input crank and others. By including the velocity specification, the designer has considerably more control of the output motion. The method of solution developed in this work uses the complex number method of mechanism synthesis. A numerical example is included.

Design of fully rotatable, roller-crank-driven, cam mechanisms for arbitrary motion specifications

The design of a non-traditional cam and roller-follower mechanism is described here. In this mechanism, the roller-crank rather than the cam is used as the continuous input member, while both complete a full rotation in each revolution and remain in contact throughout. It is noted that in order to have the cam fully rotate for every full rotation of the roller-crank, the cam cannot be a closed profile, rather the roller traverses the open cam profile twice in each cycle. Using kinematic analysis, the angular velocity of the cam when the roller traverses the cam profile in one direction, is related to the angular velocity of the cam when the roller retraces its path on the cam in the other direction. Thus, one can specify any arbitrary function relating the motion of the cam to the motion of the roller-crank for only 180 degrees of rotation in the angular velocity space. The motion of the cam in the remaining portion is then automatically determined. In specifying the arbitrary motion, many desirable characteristics such as multiple dwells, low acceleration and jerk, etc., can be obtained. Useful design equations are derived for this purpose. Using the kinematic inversion technique, the cam profile is readily obtained once the motion is specified in the angular velocity space. The only limitation to the arbitrary motion specification is making sure that the transmission angle never gets too low, so that the force will be transmitted efficiently from roller to cam. This is addressed by incorporating a transmission index into the motion specification in the synthesis process. Consequently, in this method we can specify any arbitrary motion within a permissible rone, such that the transmission index is higher than the specified minimum value. Single-dwell, double-dwell and a long hesitation motion are used as examples to demonstrate the ffectiveness of the design method. Force closure using an optimally located spring and quasi-kinetostatic analysis are also discussed. (C) 2001 Elsevier Science Ltd. All rights reserved.

Formal methods for control synthesis in partially observed environments : application to autonomous robotic manipulation

Modern robots are increasingly expected to function in uncertain and dynamically challenging environments, often in proximity with humans. In addition, wide scale adoption of robots requires on-the-fly adaptability of software for diverse application. These requirements strongly suggest the need to adopt formal representations of high level goals and safety specifications, especially as temporal logic formulas. This approach allows for the use of formal verification techniques for controller synthesis that can give guarantees for safety and performance. Robots operating in unstructured environments also face limited sensing capability. Correctly inferring a robot's progress toward high level goal can be challenging.

This thesis develops new algorithms for synthesizing discrete controllers in partially known environments under specifications represented as linear temporal logic (LTL) formulas. It is inspired by recent developments in finite abstraction techniques for hybrid systems and motion planning problems. The robot and its environment is assumed to have a finite abstraction as a Partially Observable Markov Decision Process (POMDP), which is a powerful model class capable of representing a wide variety of problems. However, synthesizing controllers that satisfy LTL goals over POMDPs is a challenging problem which has received only limited attention.

This thesis proposes tractable, approximate algorithms for the control synthesis problem using Finite State Controllers (FSCs). The use of FSCs to control finite POMDPs allows for the closed system to be analyzed as finite global Markov chain. The thesis explicitly shows how transient and steady state behavior of the global Markov chains can be related to two different criteria with respect to satisfaction of LTL formulas. First, the maximization of the probability of LTL satisfaction is related to an optimization problem over a parametrization of the FSC. Analytic computation of gradients are derived which allows the use of first order optimization techniques.

The second criterion encourages rapid and frequent visits to a restricted set of states over infinite executions. It is formulated as a constrained optimization problem with a discounted long term reward objective by the novel utilization of a fundamental equation for Markov chains - the Poisson equation. A new constrained policy iteration technique is proposed to solve the resulting dynamic program, which also provides a way to escape local maxima.

The algorithms proposed in the thesis are applied to the task planning and execution challenges faced during the DARPA Autonomous Robotic Manipulation - Software challenge.

Incremental Control Synthesis for Robotics in the Presence of Temporal Logic Specifications

This thesis presents methods for incrementally constructing controllers in the presence of uncertainty and nonlinear dynamics. The basic setting is motion planning subject to temporal logic specifications. Broadly, two categories of problems are treated. The first is reactive formal synthesis when so-called discrete abstractions are available. The fragment of linear-time temporal logic (LTL) known as GR(1) is used to express assumptions about an adversarial environment and requirements of the controller. Two problems of changes to a specification are posed that concern the two major aspects of GR(1): safety and liveness. Algorithms providing incremental updates to strategies are presented as solutions. In support of these, an annotation of strategies is developed that facilitates repeated modifications. A variety of properties are proven about it, including necessity of existence and sufficiency for a strategy to be winning. The second category of problems considered is non-reactive (open-loop) synthesis in the absence of a discrete abstraction. Instead, the presented stochastic optimization methods directly construct a control input sequence that achieves low cost and satisfies a LTL formula. Several relaxations are considered as heuristics to address the rarity of sampling trajectories that satisfy an LTL formula and demonstrated to improve convergence rates for Dubins car and single-integrators subject to a recurrence task.

Structural Synthesis of 3-DoF Spatial Fully Parallel Manipulators

In this paper, the architectures of three degrees of freedom (3-DoF) spatial, fully parallel manipulators (PMs), whose limbs are structurally identical, are obtained systematically. To do this, the methodology followed makes use of the concepts of the displacement group theory of rigid body motion. This theory works with so-called 'motion generators'. That is, every limb is a kinematic chain that produces a certain type of displacement in the mobile platform or end-effector. The laws of group algebra will determine the actual motion pattern of the end-effector. The structural synthesis is a combinatorial process of different kinematic chains' topologies employed in order to get all of the 3-DoF motion pattern possibilities in the end-effector of the fully parallel manipulator.

An approach to functional synthesis of solutions in mechanical conceptual design. Part III: Spatial configuration

These three papers describe an approach to the synthesis of solutions to a class of mechanical design problems; these involve transmission and transformation of mechanical forces and motion, and can be described by a set of inputs and outputs. The approach involves (1) identifying a set of primary functional elements and rules of combining them, and (2) developing appropriate representations and reasoning procedures for synthesising solution concepts using these elements and their combination rules; these synthesis procedures can produce an exhaustive set of solution concepts, in terms of their topological as well as spatial configurations, to a given design problem. This paper (Part III) describes a constraint propagation procedure which, using a knowledge base of spatial information about a set of primary functional elements, can produce possible spatial configurations of solution concepts generated in Part II.