Scene-relative Object Movement Research Articles

Detection of scene-relative object movement and optic flow parsing across the adult lifespan.

Moving around safely relies critically on our ability to detect object movement. This is made difficult because retinal motion can arise from object movement or our own movement. Here we investigate ability to detect scene-relative object movement using a neural mechanism called optic flow parsing. This mechanism acts to subtract retinal motion caused by self-movement. Because older observers exhibit marked changes in visual motion processing, we consider performance across a broad age range (N = 30, range: 20–76 years). In Experiment 1 we measured thresholds for reliably discriminating the scene-relative movement direction of a probe presented among three-dimensional objects moving onscreen to simulate observer movement. Performance in this task did not correlate with age, suggesting that ability to detect scene-relative object movement from retinal information is preserved in ageing. In Experiment 2 we investigated changes in the underlying optic flow parsing mechanism that supports this ability, using a well-established task that measures the magnitude of globally subtracted optic flow. We found strong evidence for a positive correlation between age and global flow subtraction. These data suggest that the ability to identify object movement during self-movement from visual information is preserved in ageing, but that there are changes in the flow parsing mechanism that underpins this ability. We suggest that these changes reflect compensatory processing required to counteract other impairments in the ageing visual system.

Ability to identify scene-relative object movement is not limited by, or yoked to, ability to perceive heading.

During locomotion humans can judge where they are heading relative to the scene and the movement of objects within the scene. Both judgments rely on identifying global components of optic flow. What is the relationship between the perception of heading, and the identification of object movement during self-movement? Do they rely on a shared mechanism? One way to address these questions is to compare performance on the two tasks. We designed stimuli that allowed direct comparison of the precision of heading and object movement judgments. Across a series of experiments, we found the precision was typically higher when judging scene-relative object movement than when judging heading. We also found that manipulations of the content of the visual scene can change the relative precision of the two judgments. These results demonstrate that the ability to judge scene-relative object movement during self-movement is not limited by, or yoked to, the ability to judge the direction of self-movement.

The Primary Role of Flow Processing in the Identification of Scene-Relative Object Movement.

Retinal image motion could be due to the movement of the observer through space or an object relative to the scene. Optic flow, form, and change of position cues all provide information that could be used to separate out retinal motion due to object movement from retinal motion due to observer movement. In Experiment 1, we used a minimal display to examine the contribution of optic flow and form cues. Human participants indicated the direction of movement of a probe object presented against a background of radially moving pairs of dots. By independently controlling the orientation of each dot pair, we were able to put flow cues to self-movement direction (the point from which all the motion radiated) and form cues to self-movement direction (the point toward which all the dot pairs were oriented) in conflict. We found that only flow cues influenced perceived probe movement. In Experiment 2, we switched to a rich stereo display composed of 3D objects to examine the contribution of flow and position cues. We moved the scene objects to simulate a lateral translation and counter-rotation of gaze. By changing the polarity of the scene objects (from light to dark and vice versa) between frames, we placed flow cues to self-movement direction in opposition to change of position cues. We found that again flow cues dominated the perceived probe movement relative to the scene. Together, these experiments indicate the neural network that processes optic flow has a primary role in the identification of scene-relative object movement.SIGNIFICANCE STATEMENT Motion of an object in the retinal image indicates relative movement between the observer and the object, but it does not indicate its cause: movement of an object in the scene; movement of the observer; or both. To isolate retinal motion due to movement of a scene object, the brain must parse out the retinal motion due to movement of the eye (“flow parsing”). Optic flow, form, and position cues all have potential roles in this process. We pitted the cues against each other and assessed their influence. We found that flow parsing relies on optic flow alone. These results indicate the primary role of the neural network that processes optic flow in the identification of scene-relative object movement.

During self-movement humans are better at judging whether an object is moving (flow parsing) than whether they will hit it (heading).

During locomotion we can use information in the retinal flow field to judge whether we will pass to the left or right of an object in the scene (heading). We can also use information in retinal flow to judge whether an object is moving relative to the scene (flow parsing). Both judgements rely on the brain identifying optic flow (global patterns of retinal motion that are characteristic of self-movement). How does the precision of these two judgments compare? Differences or similarities in precision may provide some insight into the underpinning mechanisms. We designed stimuli that allowed direct comparison of the precision of the two judgements. In the heading task, we simulated forward movement of the observer by moving a rigid scene containing wireframe objects, plus a coloured target sphere, towards the observer. Observers judged whether they would pass to the left or right of the sphere. Thresholds were measured as a function of the angle between the sphere and the simulated direction of movement. In the object movement task, the background scene and the target sphere moved on slightly different trajectories which led to a perception of scene-relative object movement. Observers judged whether the sphere was moving to the left or right in the scene. Thresholds were measured a function of the difference in direction of the background and target trajectories. Precision was significantly higher when judging object movement than when judging self-movement. Control experiments removed local cues and confirmed that the results were unchanged. In a simulated rotation experiment we found that performance on the object movement task was unaffected while judgements of self-movement were impaired. These results are compatible with the hypothesis that flow parsing is not dependent upon a prior estimate of heading; flow-parsing is either an independent process, or it precedes estimates of heading. Meeting abstract presented at VSS 2015. (Less)

Spared ability to perceive direction of locomotor heading and scene-relative object movement despite inability to perceive relative motion.

BackgroundAll contemporary models of perception of locomotor heading from optic flow (the characteristic patterns of retinal motion that result from self-movement) begin with relative motion. Therefore it would be expected that an impairment on perception of relative motion should impact on the ability to judge heading and other 3D motion tasks.Material/MethodsWe report two patients with occipital lobe lesions whom we tested on a battery of motion tasks. Patients were impaired on all tests that involved relative motion in plane (motion discontinuity, form from differences in motion direction or speed). Despite this they retained the ability to judge their direction of heading relative to a target. A potential confound is that observers can derive information about heading from scale changes bypassing the need to use optic flow. Therefore we ran further experiments in which we isolated optic flow and scale change.ResultsPatients’ performance was in normal ranges on both tests. The finding that ability to perceive heading can be retained despite an impairment on ability to judge relative motion questions the assumption that heading perception proceeds from initial processing of relative motion. Furthermore, on a collision detection task, SS and SR’s performance was significantly better for simulated forward movement of the observer in the 3D scene, than for the static observer. This suggests that in spite of severe deficits on relative motion in the frontoparlel (xy) plane, information from self-motion helped identification objects moving along an intercept 3D relative motion trajectory.ConclusionsThis result suggests a potential use of a flow parsing strategy to detect in a 3D world the trajectory of moving objects when the observer is moving forward. These results have implications for developing rehabilitation strategies for deficits in visually guided navigation.

Flow parsing and heading perception show similar dependence on quality and quantity of optic flow

Here we examine the relationship between the perception of heading and flow parsing. In a companion study we have investigated the pattern of dependence of human heading estimation on the quantity (amount of dots per frame) and quality (amount of directional noise) of motion information in an optic flow field. In the present study we investigated whether the flow parsing mechanism, which is thought to aid in the assessment of scene-relative object movement during observer movement, exhibits a similar pattern of dependence on these stimulus manipulations. Finding that the pattern of flow parsing effects was similar to that observed for heading thresholds would provide some evidence that these two complementary roles for optic flow processing are reliant on the same, or similar, neural computation. We found that the pattern of flow parsing effects observed does indeed display a striking similarity to the heading thresholds. As with judgements of heading, there is a critical value of around 25 dots per frame; below this value flow parsing effects rapidly deteriorate and above this value flow parsing effects are stable [see Warren et al. (1988) for similar results for heading]. Also, as with judgements of heading, when there were 50 or more dots there was a systematic effect of noise on the magnitude of the flow parsing effect. These results are discussed in the context of different possible schemes of flow processing to support both heading and flow parsing mechanisms.

Does optic flow parsing depend on prior estimation of heading?

We have recently suggested that neural flow parsing mechanisms act to subtract global optic flow consistent with observer movement to aid in detecting and assessing scene-relative object movement. Here, we examine whether flow parsing can occur independently from heading estimation. To address this question we used stimuli comprising two superimposed optic flow fields comprising limited lifetime dots (one planar and one radial). This stimulus gives rise to the so-called optic flow illusion (OFI) in which perceived heading is biased in the direction of the planar flow field. Observers were asked to report the perceived direction of motion of a probe object placed in the OFI stimulus. If flow parsing depends upon a prior estimate of heading then the perceived trajectory should reflect global subtraction of a field consistent with the heading experienced under the OFI. In Experiment 1 we tested this prediction directly, finding instead that the perceived trajectory was biased markedly in the direction opposite to that predicted under the OFI. In Experiment 2 we demonstrate that the results of Experiment 1 are consistent with a positively weighted vector sum of the effects seen when viewing the probe together with individual radial and planar flow fields. These results suggest that flow parsing is not necessarily dependent on prior estimation of heading direction. We discuss the implications of this finding for our understanding of the mechanisms of flow parsing.

Effects of flow field noise and density on optic flow parsing

We have previously suggested that neural flow parsing mechanisms, which subtract global optic flow consistent with observer movement, play a role in recovery of scene-relative object movement. In this study, we investigate how the quality (noise) and amount (density) of information in the optic flow field (OFF) affect flow parsing. We predict that since flow parsing involves global motion processing it should be relatively robust to noise and density manipulations. Participants fixated at the centre of a radial flow field of limited lifetime dots, and were asked to judge the direction of motion of a laterally displaced probe moving vertically upwards. To limit the contribution of local motion processing mechanisms, no flow field dots were presented in a patch of radius 3 deg centred on the probe. The 2D motion directions of the dots were subject to three levels of zero mean additive Gaussian noise (s.d. = 0, 7.5, or 15 degrees). We also varied the density of the OFF by changing the instantaneous number of dots on screen (5, 50, 100, 200 dots). We measured the relative tilt of the probe (the angular difference between the physical probe direction and the perceived direction). Analyses revealed a relatively small (although significant) decrease in the magnitude of the realative tilt effect with the increased noise (on the order of 10-20%). With respect to the density manipulation, although the relative tilt effect was significantly reduced in the lowest density condition, it was stable over all other dot densities. These data suggest that, as anticipated, flow parsing is relatively robust to variations in noise and density. The data are broadly in line with results for heading recovery when noise and density are manipulated. This provides some evidence that flow parsing and heading recovery share common mechanisms.

Does assessment of scene-relative object movement rely upon recovery of heading?

We have provided compelling evidence that the recovery of scene-relative object movement is aided by neural flow parsing mechanisms which discount (subtract) global optic flow consistent with observer movement [Warren and Rushton, 2009, Current Biology, 19, 1555?1560]. Here, we examine whether flow parsing can occur independently from heading recovery. Stimuli comprised two simultaneously presented limited lifetime optic flow fields: 1. Expanding radial flow with focus of expansion (FOE) at the centre of the display; 2. Rightwards laminar flow. This stimulus gives rise to the optic flow illusion (OFI) under which illusory shifts in heading are perceived in the direction of laminar flow. Experiment 1: Observers viewed the OFI stimulus together with a probe moving upwards (?0.6�/s) from the centre of the display. After 2 s stimulus presentation, observers reported the perceived probe motion direction by adjusting the orientation of a virtual paddle. If flow parsing depended on heading recovery under the OFI then perceived probe trajectory should tilt rightwards towards the illusory FOE. Instead, perceived motion was tilted leftwards. Experiment 2: To test that the results of experiment 1 were due to optic flow parsing we manipulated the configuration of the OFI stimulus (full field, left hemi-field only, right hemi-field only) and the start position of the probe (3 deg left/right of central fixation). When the probe was to the right, perceived trajectory tilted by 15?20� leftwards towards the centre. When the probe was to the left perceived trajectory was close to veridical. Crucially, the effects persisted when the probe was in the opposite hemi-field to the OFI stimulus - a key feature of global flow parsing. Again, these results are inconsistent with discounting of optic flow based on perceived heading under the OFI, suggesting that flow parsing does not necessarily depend on heading recovery.

Detecting object movement during self-movement: The importance of local motion contrast, position change and optic flow

We have shown that observers can identify object movement within the complex pattern of retinal motion that arises during movement of the observer. We hypothesised (Rushton & Warren, 2005) that this process exploits the brain's well-documented sensitivity to optic flow (global patterns of motion that are characteristic of self-movement) and data collected in previous studies is compatible with this hypothesis. Here we report strong direct tests of alternative solutions. We first remove the retinal disparities within the scene that are necessary for object movement to be identified by local (within disparity plane) motion contrast; we do this by presenting the scene to a single eye, whilst presenting the probe object (which varies in depth) to both eyes. We find the observers still demonstrate behaviour compatible with the identification of object movement. In a second study, we investigate the effect of removing motion (flow) from the display. We do this by presenting either a natural continuous change in (simulated) observer viewpoint, or an unnatural step change. We find that in the latter case, despite local motion and position change information being preserved, performance is significantly impaired. Taken together these results point to a central role for optic flow processing in the identification of scene-relative object movement.

Optic Flow Processing for the Assessment of Object Movement during Ego Movement

The vast majority of research on optic flow (retinal motion arising because of observer movement) has focused on its use in heading recovery and guidance of locomotion. Here we demonstrate that optic flow processing has an important role in the detection and estimation of scene-relative object movement during self movement. To do this, the brain identifies and globally discounts (i.e., subtracts) optic flow patterns across the visual scene-a process called flow parsing. Remaining motion can then be attributed to other objects in the scene. In two experiments, stationary observers viewed radial expansion flow fields and a moving probe at various onscreen locations. Consistent with global discounting, perceived probe motion had a significant component toward the center of the display and the magnitude of this component increased with probe eccentricity. The contribution of local motion processing to this effect was small compared to that of global processing (experiment 1). Furthermore, global discounting was clearly implicated because these effects persisted even when all the flow in the hemifield containing the probe was removed (experiment 2). Global processing of optic flow information is shown to play a fundamental role in the recovery of object movement during ego movement.

Perception of scene-relative object movement: Optic flow parsing and the contribution of monocular depth cues

We have recently suggested that the brain uses its sensitivity to optic flow in order to parse retinal motion into components arising due to self and object movement (e.g. Rushton, S. K., & Warren, P. A. (2005). Moving observers, 3D relative motion and the detection of object movement. Current Biology, 15, R542–R543). Here, we explore whether stereo disparity is necessary for flow parsing or whether other sources of depth information, which could theoretically constrain flow-field interpretation, are sufficient. Stationary observers viewed large field of view stimuli containing textured cubes, moving in a manner that was consistent with a complex observer movement through a stationary scene. Observers made speeded responses to report the perceived direction of movement of a probe object presented at different depths in the scene. Across conditions we varied the presence or absence of different binocular and monocular cues to depth order. In line with previous studies, results consistent with flow parsing (in terms of both perceived direction and response time) were found in the condition in which motion parallax and stereoscopic disparity were present. Observers were poorer at judging object movement when depth order was specified by parallax alone. However, as more monocular depth cues were added to the stimulus the results approached those found when the scene contained stereoscopic cues. We conclude that both monocular and binocular static depth information contribute to flow parsing. These findings are discussed in the context of potential architectures for a model of the flow parsing mechanism.

Capture of attention by scene-relative object movement during self-movement

If the observer is not moving object movement is uniquely signalled by retinal image motion. Under such circumstances a moving object captures attention. During self-movement the global image motion that results from movement of the eyes may mask the image motion that signals object movement. We investigated whether a moving object captures attention in this case. Through stereo glasses, observers viewed 24 objects arranged in a 3-D volume. Four of the objects were potential target objects that could change shape. During viewing the camera viewpoint was translated and rotated to simulate self-movement. Observers were required to identify a shape change as rapidly as possible. On every 35th trial an object within the scene would unexpectedly move at approximately the same time as the shape change. When the object that changed shape was also the object that moved, response time was shorter that when a different object moved; this is the hallmark of attentional capture. Therefore we conclude that during self-movement, unexpected object movement captures attention.

The pop out of scene-relative object movement against retinal motion due to self-movement

An object that moves is spotted almost effortlessly; it “pops out”. When the observer is stationary, a moving object is uniquely identified by retinal motion. This is not so when the observer is also moving; as the eye travels through space all scene objects change position relative to the eye producing a complicated field of retinal motion. Without the unique identifier of retinal motion an object moving relative to the scene should be difficult to locate. Using a search task, we investigated this proposition. Computer-rendered objects were moved and transformed in a manner consistent with movement of the observer. Despite the complex pattern of retinal motion, objects moving relative to the scene were found to pop out. We suggest the brain uses its sensitivity to optic flow to “stabilise” the scene, allowing the scene-relative movement of an object to be identified.