Motion Tracker: Camera-Based Monitoring of Bodily Movements Using Motion Silhouettes

Researchers in the cognitive and affective sciences investigate how thoughts and feelings are reflected in the bodily response systems including peripheral physiology, facial features, and body movements. One specific question along this line of research is how cognition and affect are manifested in the dynamics of general body movements. Progress in this area can be accelerated by inexpensive, non-intrusive, portable, scalable, and easy to calibrate movement tracking systems. Towards this end, this paper presents and validates Motion Tracker, a simple yet effective software program that uses established computer vision techniques to estimate the amount a person moves from a video of the person engaged in a task (available for download from http://jakory.com/motion-tracker/). The system works with any commercially available camera and with existing videos, thereby affording inexpensive, non-intrusive, and potentially portable and scalable estimation of body movement. Strong between-subject correlations were obtained between Motion Tracker’s estimates of movement and body movements recorded from the seat (r =.720) and back (r = .695 for participants with higher back movement) of a chair affixed with pressure-sensors while completing a 32-minute computerized task (Study 1). Within-subject cross-correlations were also strong for both the seat (r =.606) and back (r = .507). In Study 2, between-subject correlations between Motion Tracker’s movement estimates and movements recorded from an accelerometer worn on the wrist were also strong (rs = .801, .679, and .681) while people performed three brief actions (e.g., waving). Finally, in Study 3 the within-subject cross-correlation was high (r = .855) when Motion Tracker’s estimates were correlated with the movement of a person’s head as tracked with a Kinect while the person was seated at a desk (Study 3). Best-practice recommendations, limitations, and planned extensions of the system are discussed.     

Developing Motion Picture Appreciation

Abstract

A well-dressed art viewer (who appears to be affluent and highly educated) is observed gazing at several contemporary works in an art museum gallery. Out of this gentleman's view, in an adjacent gallery, a blue-collar worker enters carrying a new and attractive, but ordinary-looking door that he quickly but carefully places on the floor, leaning it against a wall near several artworks. The worker exits the gallery just as the art viewer enters it. With great concentration, the viewer contemplates all of the artworks he sees. His attention is immediately drawn to the door left behind by the workman. The viewer is intrigued by the door, which he obviously mistakes for a work of art. The door is studied intensively and at length by the viewer who seems to be searching for layers of meaning. That deep intellectual pursuit is interrupted abruptly, however, when the worker reenters the gallery, picks up the door, and carries it away. Evidently, the worker's job is to install the door elsewhere in the museum, no doubt where it will not be mistaken for a work of art. Had he been asked to comment on the art viewer's encounter with the door, the worker would surely have expressed some degree of puzzlement and criticism because—from the worker's viewpoint—anyone with even a small measure of common sense (something the “expert” art viewer apparently lacks) knows the difference between a door and a work of art, no matter how finely crafted the door may be.     

Performance Characterization of Watson Ahumada Motion Detector Using Random Dot Rotary Motion Stimuli

The performance of Watson & Ahumada''s model of human visual motion sensing is compared against human psychophysical performance. The stimulus consists of random dots undergoing rotary motion, displayed in a circular annulus. The model matches psychophysical observer performance with respect to most parameters. It is able to replicate some key psychophysical findings such as invariance of observer performance to dot density in the display, and decrease of observer performance with frame duration of the display.Associated with the concept of rotary motion is the notion of a center about which rotation occurs. One might think that for accurate estimation of rotary motion in the display, this center must be accurately known. A simple vector analysis reveals that this need not be the case. Numerical simulations confirm this result, and may explain the position invariance of MST(d) cells. Position invariance is the experimental finding that rotary motion sensitive cells are insensitive to where in their receptive field rotation occurs.When all the dots in the display are randomly drawn from a uniform distribution, illusory rotary motion is perceived. This case was investigated by Rose & Blake previously, who termed the illusory rotary motion the omega effect. Two important experimental findings are reported concerning this effect. First, although the display of random dots evokes perception of rotary motion, the direction of motion perceived does not depend on what dot pattern is shown. Second, the time interval between spontaneous flips in perceived direction is lognormally distributed (mode≈2 s). These findings suggest the omega effect fits in the category of a typical bistable illusion, and therefore the processes that give rise to this illusion may be the same processes that underlie much of other bistable phenomenon.     

Effects of Crowding and Attention on High-Levels of Motion Processing and Motion Adaptation

The motion after-effect (MAE) persists in crowding conditions, i.e., when the adaptation direction cannot be reliably perceived. The MAE originating from complex moving patterns spreads into non-adapted sectors of a multi-sector adapting display (i.e., phantom MAE). In the present study we used global rotating patterns to measure the strength of the conventional and phantom MAEs in crowded and non-crowded conditions, and when attention was directed to the adapting stimulus and when it was diverted away from the adapting stimulus. The results show that: (i) the phantom MAE is weaker than the conventional MAE, for both non-crowded and crowded conditions, and when attention was focused on the adapting stimulus and when it was diverted from it, (ii) conventional and phantom MAEs in the crowded condition are weaker than in the non-crowded condition. Analysis conducted to assess the effect of crowding on high-level of motion adaptation suggests that crowding is likely to affect the awareness of the adapting stimulus rather than degrading its sensory representation, (iii) for high-level of motion processing the attentional manipulation does not affect the strength of either conventional or phantom MAEs, neither in the non-crowded nor in the crowded conditions. These results suggest that high-level MAEs do not depend on attention and that at high-level of motion adaptation the effects of crowding are not modulated by attention.     

Integration of Motion Responses Underlying Directional Motion Anisotropy in Human Early Visual Cortical Areas

Recent imaging studies have reported directional motion biases in human visual cortex when perceiving moving random dot patterns. It has been hypothesized that these biases occur as a result of the integration of motion detector activation along the path of motion in visual cortex. In this study we investigate the nature of such motion integration with functional MRI (fMRI) using different motion stimuli. Three types of moving random dot stimuli were presented, showing either coherent motion, motion with spatial decorrelations or motion with temporal decorrelations. The results from the coherent motion stimulus reproduced the centripetal and centrifugal directional motion biases in V1, V2 and V3 as previously reported. The temporally decorrelated motion stimulus resulted in both centripetal and centrifugal biases similar to coherent motion. In contrast, the spatially decorrelated motion stimulus resulted in small directional motion biases that were only present in parts of visual cortex coding for higher eccentricities of the visual field. In combination with previous results, these findings indicate that biased motion responses in early visual cortical areas most likely depend on the spatial integration of a simultaneously activated motion detector chain.     

Enhancement of Glossiness Perception by Retinal-Image Motion: Additional Effect of Head-Yoked Motion Parallax

It has been argued that when an observer moves, a contingent retinal-image motion of a stimulus would strengthen the perceived glossiness. This would be attributed to the veridical perception of three-dimensional structure by motion parallax. However, it has not been investigated whether the effect of motion parallax is more than that of retinal-image motion of the stimulus. Using a magnitude estimation method, we examine in this paper whether cross-modal coordination of the stimulus change and the observer''s motion (i.e., motion parallax) is essential or the retinal-image motion alone is sufficient for enhancing the perceived glossiness. Our data show that a retinal-image motion simulating motion parallax without head motion strengthened the perceived glossiness but that its effect was weaker than that of motion parallax with head motion. These results suggest the existence of an additional effect of the cross-modal coordination between vision and proprioception on glossiness perception. That is, motion parallax enhances the perception of glossiness, in addition to retinal-image motions of specular surfaces.     

Periodic Tail Motion Linked to Wing Motion Affects the Longitudinal Stability of Ornithopter Flight

During slow level flight of a pigeon,a caudal muscle involved in tail movement,the levator caudae pars vertebralis,is activated at a particular phase with the pectoralis wing muscle.Inspired by mechanisms for the control of stability in flying animals,especially the role of the tail in avian flight,we investigated how periodic tail motion linked to motion of the wings affects the longitudinal stability of omithopter flight.This was achieved by using an integrative ornithopter flight simulator that included aeroelastic behaviour of the flexible wings and tail.Trim flight trajectories of the simulated ornithopter model were calculated by time integration of the nonlinear equations of a flexible multi-body dynamics coupled with a semi-empirical flapping-wing and tail aerodynamic models.The unique trim flight characteristics of ornithopter,Limit-Cycle Oscillation,were found under the sets of wingbeat frequency and tail elevation angle,and the appropriate phase angle of tail motion was determined by parameter studies minimizing the amplitude of the oscillations.The numerical simulation results show that tail actuation synchronized with wing motion suppresses the oscillation of body pitch angle over a wide range of wingbeat frequencies.