Binocular rivalry is the alternating perception that occurs when incompatible stimuli are presented to the two eyes: one monocular stimulus dominates vision and then the other stimulus dominates, with a perceptual switch occurring every few seconds. There is a need for a binocular rivalry model that accounts for both well-established results on the timing of dominance intervals and for more recent evidence on the distributed neural processing of rivalry. The model for binocular rivalry developed here consists of four parallel visual channels, two driven by the left eye and two by the right. Each channel consists of several consecutive processing stages representing successively higher cortical levels, with mutual inhibition between the channels at each stage. All stages are architecturally identical. With n the number of stages, the model is implemented as 4n nonlinear differential equations using a total of eight parameters. Despite the simplicity of its architecture, the model accounts for a variety of experimental observations: 1) the increasing depth of rivalry at higher cortical areas, as shown in electrophysiological, imaging, and psychophysical experiments; 2) the unimodal probability density of dominance durations, where the mode is less than the mean; 3) the lack of correlation between successive dominance durations; 4) the effect of interocular stimulus differences on dominance duration; and 5) eye suppression, as opposed to feature suppression. The model is potentially applicable to issues of visual processing more general than binocular rivalry.

Spatiotemporal frequency responses were measured at different levels of light adaptation for cat X and Y retinal ganglion cells. Stationary sinusoidal luminance gratings whose contrast was modulated sinusoidally in time or drifting gratings were used as stimuli . Under photopic illumination, when the spatial frequency was held constant at or above its optimum value, an X cell's responsivity was essentially constant as the temporal frequency was changed from 1 .5 to 30 Hz. At lower temporal frequencies, responsivity rolled off gradually, and at higher ones it rolled off rapidly. In contrast, when the spatial frequency was held constant at a low value, an X cell's responsivity increased continuously with temporal frequency from a very low value at 0.1 Hz to substantial values at temporal frequencies higher than 30 Hz, from which responsivity rolled off again . Thus, 0 cycles -deg' became the optimal spatial frequency above 30 Hz. For Y cells under photopic illumination, the spatiotemporal interaction was even more complex. When the spatial frequency was held constant at or above its optimal value, the temporal frequency range over which responsivity was constant was shorter than that of X cells . At lower spatial frequencies, this range was not appreciably different . As for X cells, 0 cycles deg ' was the optimal spatial frequency above 30 Hz . Temporal resolution (defined as the high temporal frequency at which responsivity had fallen to 10 impulses -s') for a uniform field was^-95 Hz for X cells and^-120 Hz for Y cells under photopic illumination . Temporal resolution was lower at lower adaptation levels. The results were interpreted in terms of a Gaussian centersurround model . For X cells, the surround and center strengths were nearly equal at low and moderate temporal frequencies, but the surround strength exceeded the center strength above 30 Hz . Thus, the response to a spatially uniform stimulus at high temporal frequencies was dominated by the surround. In addition, at temporal frequencies above 30 Hz, the center radius increased .

Binocular rivalry refers to the alternating perception that occurs when the two eyes are presented with incompatible stimuli: one monocular image is seen exclusively for several seconds before disappearing as the other image comes into view. The unseen stimulus is physically present but is not perceived because the sensory signals it elicits are suppressed. The neural site of this binocular rivalry suppression is a source of continuing controversy. We psychophysically tested human subjects, using test probes designed to selectively activate the visual system at a variety of processing stages. The results, which apply to both form and motion judgements, show that the sensitivity loss during suppression increases as the subject's task becomes more sophisticated. We conclude that binocular rivalry suppression is present at a number of stages along two visual cortical pathways, and that suppression deepens as the visual signal progresses along these pathways.

SUMMARY1. A mechanoreceptor model, developed in the preceding paper (Freeman & Johnson, 1982), was used to study the effects of vibratory intensity and frequency on the responses of slowly adapting, rapidly adapting and Pacinian afferents in monkey hairless skin. As in the previous paper almost all of the response properties studied here were accounted for by the equivalent circuit model; changes in membrane time constant and amplitude sensitivity accounted for the differences between the three mechanoreceptive fibre types.2. The stimulus-response function of primary concern was the relationship between impulse rate and vibratory amplitude. This relationship had the same general form in each of the three fibre types. Amplitudes, I, less than Io produced no impulse on any stimulus cycles. Amplitudes greater than I, produced one impulse on every cycle. As I rose from Io to I, the impulse rate rose monotonically from 0 to 1 impulse/cycle. For each fibre type the form ofthis ramp depended on the stimulus frequency. A. W. FREEMAN AND K. 0. JOHNSON 6. The model's membrane time constant was adjusted to match the observed reduction in the (IO, I,) slope with increasing stimulus frequency. The time constants required for least-squares fitting were 58, 29 and 4-2 msec for slowly adapting, rapidly adapting and Pacinian afferents, respectively; these values are of the same order as those obtained in the preceding paper.7. Receptor sensitivity varied across the frequency spectrum, slow adaptors being most sensitive at low frequencies, rapidly adapting units at mid-range, and Pacinians at the high frequencies. According to the model, the high frequency roll-off in a receptor's tuning curve is due to the current integrating properties of receptor membrane, and the low frequency roll-off is due to a high pass filter, presumably mechanical, situated in the tissues between the stimulus probe and receptor membrane.8. Impulse phase advances with increasing stimulus intensity in both receptor and model. The ability of the model to fit both the rate-intensity function and phase advance functions in individual receptors is demonstrated.