The contribution of flow and extra-retinal signals to control of steering.
R.M. Wilkie & J.P. Wann.
Dept Psychology, Univ Reading, Berks, UK.
Purpose: A considerable volume of research has been directed
towards investigating the problem of perceiving heading when locomoting along
linear paths. In many real world settings, however, we look where we wish to go,
choose a path, and change our direction of motion to realise that goal or path.
Following on from Wann J.P. & Wilkie R.M. ARVO 2001, we used the paradigm of
independently adjusting retinal flow and gaze motion to explore their roles in
actively steering towards a target along a smooth trajectory.
Methods: Three main sources of information were manipulated:
extra-retinal signals (ER), retinal flow (RF) and visual frames of reference
(VF) in the form of dark frame edges that provide a static reference for target
motion. We compared pure RF (simulated rotation), with RF + ER, and RF + ER +
VF. In all cases RF, ER and VF motion signals were based on the instantaneous
steering response of participants. We also introduced three different ground
textures providing high, medium and very low quality optic flow.
Results: The gaze motion information, arising from either ER or
VF, ensured accurate steering for all textures. Removal of VF (while maintaining
appropriate ER information) made no difference for the high quality flow
condition, which remained accurate, but significantly increased errors for the
medium and low quality flow. Removal of all gaze motion information (simulated
rotation) dramatically increased steering errors and variability for all
textures, though quality of flow still predicted magnitude of error. This
supports our previous findings with passive heading detection, whilst
highlighting the importance of gaze motion information in real world steering
tasks. It also demonstrates that different sources of information (RF, ER, VF)
are combined for accurate perception and action in a locomotor context (Warren
W.H. et al. Invest. Opthalmol. Vis. Sci. 41, S812).
The contribution of flow and extra-retinal signals to judgements of
heading.
J.P. Wann & R.M. Wilkie.
Department of Psychology, University of Reading, Reading, UK.
Purpose: Do we require extra-retinal information about gaze motion
to perceive heading from retinal flow, when gaze is eccentric to your path? A
series of studies by Banks et al (1996 Vis. Res. 36, 431-443) demonstrated a
systematic bias if the ratio of actual/simulated gaze rotation was reduced. This
supported the role of extra-retinal information, but because gaze pursuit was
used to introduce rotation into retinal flow, this was not a strong test of the
role of gaze pursuit in resolving rotation and translation components in retinal
flow.
Methods: We revisited the theme of heading estimation using a
paradigm where the extra-retinal information arising from gaze motion was
manipulated independently from the simulated rotation produced within the optic
flow field: By moving the viewport around the stimuli (as if moving the screen),
extra-retinal signals were induced without changing the retinal image.
Results: Observer performance using real eye movements was
compared to that of simulated retinal flow, with or without extra-retinal
signals. In all three cases the retinal flow pattern was equivalent. Heading
judgements with appropriate eye movements (extra-retinal signals) proved more
accurate than those with only retinal information and those with inappropriate
(reverse direction) extra-retinal information produced large errors. These
results support Banks et al by using independent adjustment of the contributory
variables (flow and gaze motion). This paradigm was then directly translated to
the task of steering to a point of fixation (see Wilkie R.M. & Wann J.P.
ARVO 2001).
Examples of Displays
Go to animated
displays page >>
