Observation and analysis of large-scale human motion

Many team sports include complex human movement, which can be observed at different levels of detail. Some aspects of the athlete's motion can be studied in detail using commercially available high-speed, high-accuracy biomechanical measurement systems. However, due to their limitations, these devices are not appropriate for studying large-scale motion during a game (for example, the motion of a player running across the entire playing field). We describe an alternative approach to studying such large-scale motion, and present a video-based, computer-aided system, developed specifically for the purpose of acquiring large-scale motion data. The baseline of our approach consists of sacrificing much of the spatial accuracy and temporal resolution of widely used biomechanical measurement systems, to obtain data on human movement that span large areas and long intervals of time. Data can be obtained for each of the observed athletes with reasonable amount of operator work. The system was developed using the recordings of a handball match. Several field tests were performed to assess measurement error, including comparison to one of the widely available biomechanical measurement systems. With the help of the system presented, we could obtain position data for all 14 handball players on a 40 x 20 m large court with RMS error better than 0.6 m, covering 1 h of action. Several results, obtained during the handball match study are presented, in order to highlight the importance of large-scale motion acquisition.     

Micro motion analysis of the human arm,performing with a body powered hook

The study was designed to investigate the performance of amputees by means of a micro motion and time technique. The objectives of the investigation were to analyze the basic motion elements that construct the locomotion of the amputated arm while performing a manipulative task, to relate and compare the motions of a work cycle between amputated and natural arms for the same task, and to investigate qualitatively and quantitatively the contribution of each motion element to overall performance. Based on the results of this investigation, the researcher was able to point out which part of the motion should be considered for improving time performance of the body powered mechanical hook used by amputees.Twenty-six subjects participated in this experiment, half of whom were amputees and the other half non-amputees. The study utilized the Methods Time Measurement (MTM) technique — a widely used industrial engineering procedure — as an evaluative method.The findings showed significant differences in both motion and time elements between the performance of amputees and non-amputees — mainly due to the motion element grasp. Outlines for improvement of the design of the hook on the basis of this study are discussed.     

Motion transparency: making models of motion perception transparent

In daily life our visual system is bombarded with motion information. We see cars driving by, flocks of birds flying in the sky, clouds passing behind trees that are dancing in the wind. Vision science has a good understanding of the first stage of visual motion processing, that is, the mechanism underlying the detection of local motions. Currently, research is focused on the processes that occur beyond the first stage. At this level, local motions have to be integrated to form objects, define the boundaries between them, construct surfaces and so on. An interesting, if complicated case is known as motion transparency: the situation in which two overlapping surfaces move transparently over each other. In that case two motions have to be assigned to the same retinal location. Several researchers have tried to solve this problem from a computational point of view, using physiological and psychophysical results as a guideline. We will discuss two models: one uses the traditional idea known as ‘filter selection’ and the other a relatively new approach based on Bayesian inference. Predictions from these models are compared with our own visual behaviour and that of the neural substrates that are presumed to underlie these perceptions.     

An analysis of induced motion

The motion induced in a stationary object by a displacing surround invariably was accompanied by an apparent change in the erocentric location of that object. The degree of this apparent spatial displacement significantly correlated with the amount of induced motion seen. Induced movement was, therefore, interpreted in terms of a subjective change in the location of an object in space, as opposed to the object-relative displacement hypothesis proposed by Duncker in 1929. These findings were extended to test critically such other concepts in motion perception as separation of systems.     

External analysis with three-mode principal component models

Through external analysis of two-mode data one attempts to map the elements of one mode (e.g., attributes) as vectors in a fixed space of the elements of the other mode (e.g., stimuli). This type of analysis is extended to three-mode data, for instance, when the ratings are made by more individuals. It is described how alternating least squares algorithms for three-mode principal component analysis (PCA) are adapted to enable external analysis, and it is demonstrated that these techniques are useful for exploring differences in the individuals' mappings of the attribute vectors in the fixed stimulus space. Conditions are described under which individual differences may be ignored. External three-mode PCA is illustrated with data from a person perception experiment, designed after two studies by Rosenberg and his associates whose results were used as external information.We gratefully acknowledge the assistance of Piet Brouwer in implementing the external analysis options in the TUCKALS programs.     

Rating scale analysis with latent class models

A general approach for analyzing rating data with latent class models is described, which parallels rating models in the framework of latent trait theory. A general rating model as well as a two-parameter model with location and dispersion parameters, analogous to Andrich's Dislocmodel are derived, including parameter estimation via the EM-algorithm. Two examples illustrate the application of the models and their statisticalcontrol. Model restrictions through equality constrains are discussed and multiparameter generalizations are outlined.     

Visual motion detection models: features and frequency filters

Two kinds of models have been proposed for taking into account the sensory processes at work in the detection of visual motion: the feature model and the frequency-filter model. The problem of the complementarity of these models is raised. On the basis of empirical data, it is proposed that they are consistent.     

Confirmatory factor analysis of ITPA models with language-impaired children

The major objective of the current investigation was to provide an evaluation of the validity of the ITPA for language-impaired children using confirmatory factor analysis. In addition to an examination of the validity of the traditional ITPA models, a new model is tested. The new model is based on considerations of the structural complexity of the ITPA subtests. and allows one to interpret ITPA results in processing terms. The results suggest that four-factor models provided the best representation of the ITPA scores in language-impaired children. Compared with the traditional ITPA models, the new model showed either an improved or equal goodness of fit. Language-impaired children tested with the ITPA (n = 100) exhibited a profile pattern suggesting that subtests with low processing demands are "easier" than tests with high processing demands. Thus, the present study supports the contention that language-impaired children suffer from constraints on information processing.     

Influence analysis of structural equation models with polytomous variables

The estimation of model parameters in structural equation models with polytomous variables can be handled by several computationally efficient procedures. However, sensitivity or influence analysis of the model is not well studied. We demonstrate that the existing influence analysis methods for contingency tables or for normal theory structural equation models cannot be applied directly to structural equation models with polytomous variables; and we develop appropriate procedures based on the local influence approach of Cook (1986). The proposed procedures are computationally efficient, the necessary bits of the proposed diagnostic measures are readily available following an usual fit of the model. We consider the influence of an individual cell frequency with respect to three cases: when all parameters in an unstructured model are of interest, when the unstructured polychoric correlations are of interest, and when the structural parameters are of interest. We also consider the sensitivity of the parameters estimates. Two examples based on real data are presented for illustration.The work described in this paper was partially supported by a Chinese University of Hong Kong Postdoctoral Fellows Scheme and a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (RGC Ref. No. CUHK4186/98P). We are indebted to P.M. Bentler and M.D. Newcomb for providing the data set, Michael Leung for his assistance, and the Editor and the referees for some very valuable comments.     

Wavelet-frequency analysis for the detection of discontinuities in switched system models of human balance

This paper is concerned with detecting the presence of switching behavior in experimentally obtained posturographic data sets by means of a novel algorithm that is based on a combination of wavelet analysis and Hilbert transform. As a test-bed for the algorithm, we first use a switched model of human balance control during quiet standing with known switching behavior in four distinct configurations. We obtain a time–frequency representation of a signal generated by our model system. We are then able to detect manifestations of discontinuities (switchings) in the signal as spiking behavior. The frequency of switchings, measured by means of our algorithm and detected in our models systems, agrees with the frequency of spiking behavior found in the experimentally obtained posturographic data.     

Intersections of behavior analysis with cognitive models of contingency detection

Bower and Watson have offered, respectively, a logical hypothesis-testing model and a conditional probability model of contingency detection by young infants. Although each could represent cognitive processes concomitant with operant learning, empirical support for these models is sparse. The limitations of each model are discussed, and suggestions are made for a more parsimonious approach by focusing on the areas of overlap between the two.     

Visual motion interferes with tactile motion perception

Previous studies have demonstrated that visual apparent motion can alter the judgment of auditory apparent motion. We investigated the effect of visual apparent motion on judgments of the direction of tactile apparent motion. When visual motion was presented at the same time as, but in a direction opposite to, tactile motion, accuracy in judging the direction of tactile apparent motion was substantially reduced. This reduction in performance is referred to as 'the congruency effect'. Similar effects were observed when the visual display was placed either near to the tactile display or at some distance from the tactile display (experiment 1). In experiment 2, the relative alignment between the visual and tactile directions of motion was varied. The size of the congruency effect was similar at 0 degrees and 45 degrees alignments but much reduced at a 90 degrees alignment. In experiment 3, subjects made confidence ratings of their judgments of the direction of the tactile motion. The results indicated that the congruency effect was not due to subjects being unsure of the direction of motion and being forced to guess. In experiment 4, static visual stimuli were shown to have no effect on the judgments of direction of the tactile stimuli. The extent to which the congruency effect reflects capture effects and is the result of perceptual versus post-perceptual processes is discussed.     

The cognitive representation of an event involving human motion

Human motion is analyzed as a function of the temporal pattern of relative stability and instability of the body in relation to its supporting ground. The cognitive representation of an event unit involving human motion is described as some small number of relatively stable, preparatory motions, followed by an unstable, completing motion. Two experiments tested the coherence of the preparatory motion (s)-completing motion structure, and the visual conditions under which an unstable motion is encoded as the completing motion of an event unit. Preparatory-completing structures were recognized more easily than completing-preparatory structures. The offset of contour motion at the center of an unstable motion was found to be an important variable in the encoding of a completing motion.     

Vector analysis in visual perception of rolling motion

Summary The well-known visual effect of seeing an isolated, point on the rim of a rolling wheel as moving in a cycloidal track was used for a study of visual vector analysis under non-redundant conditions.Under actual experimental conditions a mean difference of about 20% was found between the mathematically predicted and the perceptual vector analysis. A considerable interperson variation in deviation from the analysis was found. Some Ss showed a deviation of only a few per cent from the mathematically correct response.This investigation was made possible by grants from the Swedish Council for Social Science Research.     

Activation of human motion processing areas during event perception

Observers are able to segment continuous everyday activity into meaningful parts. This ability may be related to processing low-level visual cues, such as changes in motion. To address this issue, the present study combined measurement of evoked responses to event boundaries with functional identification of the extrastriate motion complex (MT+) and the frontal eye field (FEF), two regions related to motion perception and eye movements. The results provided strong evidence that MT+ is activated by event boundaries: Individuals' MT+ regions showed strong responses to event boundaries, and MT+ was collocated with a lateral posterior region that responded at event boundaries. The evidence regarding the FEF was less conclusive: The FEF showed reliable but relatively reduced responses to event boundaries, but the FEF was medial and superior to a frontal area that responded at event boundaries. These results suggest that motion cues, and possibly eye movements, may play key roles in event structure perception.     

An oscillatory correlation model of visual motion analysis

We describe and evaluate a model of motion perception based on the integration of information from two parallel pathways: a motion pathway and a luminance pathway. The motion pathway has two stages. The first stage measures and pools local motion across the input animation sequence and assigns reliability indices to these pooled measurements. The second stage groups locations on the basis of these measurements. In the luminance pathway, the input scene is segmented into regions on the basis of similarities in luminance. In a subsequent integration stage, motion and luminance segments are combined to obtain the final estimates of object motion. The neural network architecture we employ is based on LEGION (locally excitatory globally inhibitory oscillator networks), a scheme for feature binding and region labeling based on oscillatory correlation. Many aspects of the model are implemented at the neural network level, whereas others are implemented at a more abstract level. We apply this model to the computation of moving, uniformly illuminated, two-dimensional surfaces that are either opaque or transparent. Model performance replicates a number of distinctive features of human motion perception.