Human spatial orientation in the pitch dimension

Natural statistics of human head orientation constrain models of vestibular processing

Head orientation relative to gravity determines how gravity-dependent environmental structure is sampled by the visual system, as well as how gravity itself is sampled by the vestibular system. Therefore, both visual and vestibular sensory processing should be shaped by the statistics of head orientation relative to gravity. Here we report the statistics of human head orientation during unconstrained natural activities in humans for the first time, and we explore implications for models of vestibular processing. We find that the distribution of head pitch is more variable than head roll and that the head pitch distribution is asymmetrical with an over-representation of downward head pitch, consistent with ground-looking behavior. We further suggest that pitch and roll distributions can be used as empirical priors in a Bayesian framework to explain previously measured biases in perception of both roll and pitch. Gravitational and inertial acceleration stimulate the otoliths in an equivalent manner, so we also analyze the dynamics of human head orientation to better understand how knowledge of these dynamics can constrain solutions to the problem of gravitoinertial ambiguity. Gravitational acceleration dominates at low frequencies and inertial acceleration dominates at higher frequencies. The change in relative power of gravitational and inertial components as a function of frequency places empirical constraints on dynamic models of vestibular processing, including both frequency segregation and probabilistic internal model accounts. We conclude with a discussion of methodological considerations and scientific and applied domains that will benefit from continued measurement and analysis of natural head movements moving forward.

Body-relative horizontal-vertical anisotropy in human representations of traveled distances

A growing number of studies investigated anisotropies in representations of horizontal and vertical spaces. In humans, compelling evidence for such anisotropies exists for representations of multi-floor buildings. In contrast, evidence regarding open spaces is indecisive. Our study aimed at further enhancing the understanding of horizontal and vertical spatial representations in open spaces utilizing a simple traveled distance estimation paradigm. Blindfolded participants were moved along various directions in the sagittal plane. Subsequently, participants passively reproduced the traveled distance from memory. Participants performed this task in an upright and in a 30° backward-pitch orientation. The accuracy of distance estimates in the upright orientation showed a horizontal–vertical anisotropy, with higher accuracy along the horizontal axis compared with the vertical axis. The backward-pitch orientation enabled us to investigate whether this anisotropy was body or earth-centered. The accuracy patterns of the upright condition were positively correlated with the body-relative (not the earth-relative) coordinate mapping of the backward-pitch condition, suggesting a body-centered anisotropy. Overall, this is consistent with findings on motion perception. It suggests that the distance estimation sub-process of path integration is subject to horizontal–vertical anisotropy. Based on the previous studies that showed isotropy in open spaces, we speculate that real physical self-movements or categorical versus isometric encoding are crucial factors for (an)isotropies in spatial representations.

Gender Differences in Perception of Self-Orientation: Software or Hardware?

We evaluated the contribution of attentional strategy to the perception of self-orientation with and without a body tilt in the median plane. Reinking et al (1974 Journal of Personality and Social Psychology30 807–811) found that the frame dependence of females on the rod-and-frame test could be mediated by instructions prompting them to focus on internal cues (ie arising from inside of the body). Here, we measured the influence of attentional instructions on the perception of the morphological horizon. Eleven females and thirteen males estimated their morphological horizon in an upright and a 45° body tilt in the median plane under three instruction conditions. All participants first performed without attentional instructions. Then, participants performed under both internal and external attentional instructions. For females, but not for males, perception of morphological horizon was more footward in the supine than in the upright orientation. Although instructions did not eliminate gender differences, internal instructions allowed females to reduce their perceptual bias in the supine orientation.

Estimation of the horizon in photographed outdoor scenes by human and machine

We present three experiments on horizon estimation. In Experiment 1 we verify the human ability to estimate the horizon in static images from only visual input. Estimates are given without time constraints with emphasis on precision. The resulting estimates are used as baseline to evaluate horizon estimates from early visual processes. Stimuli are presented for only 153 ms and then masked to purge visual short-term memory and enforcing estimates to rely on early processes, only. The high agreement between estimates and the lack of a training effect shows that enough information about viewpoint is extracted in the first few hundred milliseconds to make accurate horizon estimation possible. In Experiment 3 we investigate several strategies to estimate the horizon in the computer and compare human with machine "behavior" for different image manipulations and image scene types.

The perception of visually presented yaw and pitch turns: Assessing the contribution of motion, static, and cognitive cues

Terrestrial gravity restricts human locomotion to surfaces in which turns involve rotationsaround the body axis. Because observers are usually upright, one might expect the effects of gravity to induce differences in the processing of vertical versus horizontal turns. Subjects observed visual scenes of bending tunnels, either statically or dynamically, as if they were moving passively through the visual scene and were then asked to reproduce the turn deviation of the tunnel with a trackball. In order to disentangle inertia-related (earth-centered) from vision-related (body-centered) factors, the subjects were either upright or lying on their right side during the observations. Furthermore, the availability of continuous optic flow, geometrical cues, and eye movement were manipulated in three experiments. The results allowed us to characterize the factors' contributions as follows. Forward turns (pitch down) with all cues were largely overestimated, as compared with backward turns (pitch up). First, eye movements known to be irregular for vertical stimulation were largely responsible for this asymmetry. Second, geometry-based estimations are, to some extent, asymmetrical. Third, a cognitive effect corresponding to the evaluation of navigability for upward and downward turns was found (i.e.,top-down influences, such as the fear of falling often reported), which tended to increase the estimation of turns in the direction of gravity.

Contribution of action to perception of self-orientation in humans

In this study, we evaluated the effect of action on the perception of an egocentric illusion. Eighteen participants were asked to indicate the perceived morphological horizon under two backward body tilts from upright in the median plane (i.e. pitch) using five different response modes. The response modes varied in the degree of motor and cognitive involvement. Differences in perception of the morphological horizon between the two body tilts were significant only when proximal limb control was not involved. These results suggest that motor involvement and frame of reference may both be important in visual-vestibular illusions.

The reproduction of vertical and oblique orientations in the visual, haptic, and somato-vestibular systems

This study investigates whether the vertical orientation may be predominantly used as an amodal reference norm by the visual, haptic, and somato-vestibular perceptual systems to define oblique orientations. We examined this question by asking the same sighted adult subjects to reproduce, in the frontal (roll) plane, the vertical (0 degree) and six oblique orientations in three tasks involving different perceptual systems. In the visual task, the subjects adjusted a moveable rod so that it reproduced the orientation of a visual rod seen previously in a dark room. In the haptic task, the blindfolded sighted subjects scanned an oriented rod with one hand and reproduced its orientation, with the same hand, on a moveable response rod. In the somato-vestibular task, the blind-folded sighted subjects, sitting in a rotating chair, adjusted this chair in order to reproduce the tested orientation of their own body. The results showed that similar oblique effects (unsigned angular error difference between six oblique orientations and vertical orientation) were observed across the three tasks. However, there were no positive correlations between the visual, haptic, and somato-vestibular oblique effects. Moreover, in some oblique orientations, there was a tendency to overestimate the angle between the oblique orientation and the vertical orientation. This effect varied according to the orientation value and the modality. Taken together, these findings suggest that although vertical orientation is used as a reference norm in the visual, haptic, and somato-vestibular systems to define oblique orientations, specific processing mechanisms seem to be at work in each perceptual system.

Effects of supine body position and low radial accelerations on the visually perceived apparent zenith

The visually perceived eye level (VPEL) has been shown to shift toward the lower part of the body in upright subjects facing toward the axis of rotation on a centrifuge. This shift occurs in the same direction as the shift in the gravito-inertial forces (Gis) produced by very low radial acceleration (centrifugation) combined with gravity. The purpose of this study was to determine whether the same phenomenon affects the visually perceived apparent zenith (VPAZ) in subjects in a supine position. Twelve supine subjects were instructed to set a luminous target to the VPAZ, either while they were in total darkness and motionless or while undergoing very low centrifugation. Data showed that Gis induced a VPAZ shift similar to that observed for the VPEL. Thus, as is the case for the VPEL, the corresponding logarithmic psychophysical function of the VPAZ may be considered to be a type of oculogravic illusion phenomenon with differences in the subjects' that differs from subject to subject, depending on the subject's sensitivity to low radial accelerations. Data on VPEL and VPAZ support the notion that the subjective perception of eye level in total darkness takes into account changes--even if extremely slight-in the direction of the gravito-inertial forces produced by the combination of gravity and low radial accelerations, although subjects are unaware of the Gi shift. However, depending on the intensity of the radial acceleration and the angular deviation of Gi relative to G, the shift of the VPEL and the VPAZ can be either amplified or attenuated. Moreover, differences between VPEL and VPAZ responses suggest two explanatory assumptions--namely, that this is (1) a peripheral phenomenon dependent on the neurophysiological anisotropy of the otolithic system or (2) a central phenomenon dependent on the relevance assigned to the peripheral information by the integrative sensory functions and the associative processes.

Apparent body tilt and postural aftereffect

Apparent orientation of the body tilted laterally in the frontal plane was studied with the methods of absolute judgments in four experiments. In Experiment 1, 17 subjects, who maintained the normal adaptation of body to gravity, estimated their body tilts under the condition of seeing the gravitational vertical and under the condition of eliminating it. The results showed that (1) there was not a significant difference between the two conditions and (2) the small tilts of less than 45 degrees were exactly estimated, whereas the large tilts of 45 degrees-108 degrees were overestimated. In Experiment 2, 10 subjects estimated their body tilts under three velocities of a rotating chair on which each subject was placed. Although both body tilt and chair velocity were found to influence tilt estimation, the effect of body tilt was overwhelmingly greater than that of chair velocity. In Experiment 3, 11 subjects adapted their bodies to a 72 degrees left tilt for 10 min and then estimated various body tilts around the adapting tilt. The estimations obtained under the 72 degrees adaptation were lower than those obtained under the 0 degree adaptation, and this reduction was greater for the test tilt that was farther away from the adapting tilt. In Experiment 4, 11 subjects adjusted their own body tilts to designated angles. The results confirmed the outcomes of absolute estimation in Experiments 1-3. From these findings and past literature, the judgments of body tilt were considered to be subserved by a single sensory process that was based on the cutaneous and muscular proprioceptors, rather than the vestibular and joint proprioceptors.

Shift of subjective reference and visual orientation during slow pitch tilt for the seated human subject

We examined the ability to assess subjective orientation and orientation of an external visual object during pitch tilt. Subjects were seated, restrained, and in darkness in a simulator and estimated when they were 0 degree, 45 degrees, and 90 degrees forwards and backwards from upright during pitching at 1 degree/s. They temporarily stopped in these positions and set a 5 cm luminous cube, cockpit mounted at 60 cm from the nasium, to earth vertical. Estimates of subjective tilt were consistently greater than actual tilt. Overestimations were increased by preceding tilts in the opposite direction, particularly when tilting from forwards, where subjects sometimes estimated they were tilted backwards when the machine was tilted forwards. Subjects were surprised with their estimates, and reported disorientation. Regardless, settings of the visual vertical made "intuitively" were largely accurate. Subjective estimates could be construed as "accurate" if one assumes that the rostro-caudal axis of the head was referenced for estimates of upright and forwards and a trunk-leg axis for backwards. Because labyrinthine defective patients behaved as normal subjects, task performance must have been based on proprioception. The overestimation of tilt is exploited in fairground illusions and may account for the common experience when driving, that hills seem much steeper than they are.

Visual and somesthetic influences on postural orientation in the median plane

We investigated optic and somesthetic contributions to perceived body orientation in the pitch dimension. In a within-subject factorial design, each of 12 subjects attempted to set his/her body erect or 45 degrees back from erect while restrained in a movable bed surrounded by an adjustable box. The box provided a visual environment consisting of either a grid pattern, two luminous lines, or complete darkness. Both the grid pattern and the luminous lines were effective at biasing settings of body position when the box was pitched; the pitched grid was more effective than the pitched lines. Although the pitch of the box influenced orientation to both goals, the effect was greater for the diagonal goal than for the erect goal. We present a model of postural orientation in the median plane that involves vestibular, somatosensory, and visual inputs.

Effect of structured visual environments on apparent eye level

Each of 12 subjects set a binocularly viewed target to apparent eye level; the target was projected on the rear wall of an open box, the floor of which was horizontal or pitched up and down at angles of 7.5 degrees and 15 degrees. Settings of the target were systematically biased by 60% of the pitch angle when the interior of the box was illuminated, but by only 5% when the interior of the box was darkened. Within-subjects variability of the settings was less under illuminated viewing conditions than in the dark, but was independent of box pitch angle. In a second experiment, 11 subjects were tested with an illuminated pitched box, yielding biases of 53% and 49% for binocular and monocular viewing conditions, respectively. The results are discussed in terms of individual and interactive effects of optical, gravitational, and extraretinal eye-position information in determining judgements of eye level.