The distribution of image velocities on the retinas of a moving observer can be described in terms of either vector directions or vector magnitudes (i.e. speeds). Whereas gaze rotations and path curvatures distort both the directional and the speed structure of the flow field, depth irregularities in the scene affect only the pattern of image speeds without changing the pattern of motion directions. Experimentally, it is possible to manipulate the speed distribution independent of motion direction without losing the perception of depth order and flow (Schrater, Knill & Simoncelli, 2001). It thus appears that the speed structure of the flow field is informative in its own right as it reflects the three-dimensional structure of the environment while taking self-motion into account.

Introducing speed asymmetries to a radial flow with symmetrical directional structure has been shown to bias heading perception (Dyre & Andersen, 1997). It is thus of particular interest whether the speed pattern of optic flow controls certain aspects of actual ego-motion. While the human data bearing on this question is sparse, there is convincing evidence that relative image speed guides directly the flight trajectory of the honeybee (Srinivasan & Zhang, 2000). As the bee negotiates narrow gaps or flies through a tunnel it passes precisely through the center of the opening by balancing image speed in the two eyes (i.e. moving towards the side projecting slower image motion and away from the side projecting faster image motion). This centering behavior is governed by relative angular speed alone and is insensitive to large differences in the spatial structure, contrast and motion direction on the two sides. It is tempting to hypothesize that humans employ a similar speed equalization strategy to avoid obstacles along their heading path (Warren, 1998).

If such a speed equalization mechanism is really at work, the human visual system should exhibit particular sensitivity to bilateral speed asymmetries, possibly mediated by complex motion mechanisms encoding global patterns of retinal flow. Yet there is no psychophysical evidence to support this claim. It has been repeatedly shown that speed discrimination thresholds for optic flow stimuli are comparable to those obtained with simple translation (Sekuler, 1992; Clifford, Beardsley & Vaina, 1999). It was also found that speed discrimination improves to the same extent following adaptation to motion and to counterphase flicker which implicates only low-level local temporal frequency detectors in the speed adaptation / discrimination task (Clifford and Wenderoth, 1999).

We assert that the above findings might be limited to the particular experimental conditions: the temporal nature of the discrimination task, unilateral adaptation and stimuli of limited spatial extent. Since it is possible that global motion detectors are exclusively involved in long range speed comparisons we reexamined the issue with large-scale optic flow stimulation while employing bilateral simultaneous speed discrimination across space. Following adaptation to our realistic "tunnel" display we report robust enhancement of bilateral speed discrimination within the optic flow. The pattern of results is different from that obtained with spatially confined motion patches: the global post-adaptive enhancement is twice as strong and motion specific and it generalizes across a wide range of test contrasts around the adapted level. Our initial results suggest that global speed comparisons within the optic flow are at least partially mediated by motion sensitive neurons with large and overlapping receptive fields whose responses remain unaffected by suprathreshold variations of contrast.

References:

Schrater, P.R., Knill, D.C. & Simoncelli, E.P. (2001). Perceiving visual expansion without optic flow. Nature, 410(6830), 816-819.

Dyre, BP., & Andersen, G.J. (1997). Image velocity magnitudes and perception of heading. Journal of Experimental Psychology: Human Perception and Performance, 23(2), 546-565.

Srinivasan, M.V., & Zhang,S-W. (2000). Visual navigation in flying insects. In M.Lappe (Ed.). Neuronal Processing of Optic Flow, International Review of Neurobiology, 44, 67-91.

Warren, W.H. (1998). Visually controlled locomotion: 40 years later. Ecological Psychology, 10(3-4), 177-219.

Sekuler, A.B. (1992). Simple pooling of unidirectional motion predicts speed discrimination for looming stimuli. Vision Research, 32(12), 2277-2288.

Clifford, C.W.G, Beardsley, S.A. & Vaina, L.M. (1999). The perception and discrimination of speed in complex motion. Vision Research, 39, 2213-2227.

Clifford, C.W.G, & Wenderoth, P. (1999). Adaptation to temporal modulation can enhance differential speed sensitivity. Vision Research, 39. 4324-4332.