Abstract
Moving at 1.5 m/sec towards a cloud of dots that extends beyond 5 m distance, an eye rotation of 3 deg/sec displaces the centre in the retinal flow by about 10 deg. If the heading direction is represented by a set of templates tuned to different loci for the center of expansion, how does the visual system compensate for such an effect of eye rotation? One suggestion is that oculo-motor signals are used to remove the rotational component at the level of the local motion detectors, which then form the input to a map of head-centric expansion templates (Royden et al. 1994). This would predict that the perceived heading is always shifted relative to the center in the retinal flow in the direction opposite to the eye rotation. To test this idea, we presented to the moving eye (1) pure retinal expansion, or (2) pure expansion on the screen. The cloud stimuli were presented in random order, for varying horizontal heading directions (1.5 m/s) and amounts of eye rotation (up to 3 deg/sec) about the vertical axis. Subjects were asked to indicate their direction of self-motion at the end of each trial. For condition ( 1 ), perceived heading differed less than 2.5 degrees from the center of the retinal expansion. For condition (2), a nearly complete compensation for the eye rotation occurred and perceived heading differed less than 2.5 degrees from the simulated heading direction. This indicates that the oculomotor signal affects the perceived heading only if the rotational component is present in the retinal flow. These results are more consistent with an alternative model in which the oculo-motor signal is combined higher up in the visual system with motion templates, that indicate the amount of retinal rotational flow.
Original language | English |
---|---|
Pages (from-to) | S78 |
Journal | Investigative Ophthalmology and Visual Science |
Volume | 38 |
Issue number | 4 |
Publication status | Published - 1997 |