We have been working on this striking illusion that was discovered by Yoram Bonneh. There are a number of similarities to rivalry (See Carter and Pettigrew 2003  pdf).
Look steadily at one stationary point, such as one of the yellow discs, as blinks and sudden eye movements destroy the illusion Notice that one, two or all of the yellow discs will disppear and reappear.
 
 

We can make the disappearing yellow dots reappear by disrupting the activity of the left hemisphere using TMS (Funk and Pettigrew 2003  pdf). Alternatively, we can increase the duration of the disappearance by disrupting the activity of the Right hemsiphere with precisely time pulses (the Right hemisphere seems to be much more picky about the prcise timing of the TMS pulse than the Left, perhaps associated with the larger blocks of time dealt with by the Left ).
Our interpretation is that the left hemisphere is "into denial" as Ramachandran showed so nicely ( in patients suffering from right hemisphere damage,  who felt less compelled to deny their paralysis when right hemisphere function was temporarily improved by caloric stimulation).
If we are correct, the illusion and perceptual rivalry may share the same competitive tussle between the right and left sides of the brain. The disappearance of the yellow discs may thus represent the ascendancy of the Left, while their reappearance may represent the ascendancy of the Right.
Critics have pointed out that there is only one Left hemisphere while there are a number of different possible patterns of disappearance of the yellow discs (8 to be precise). This may be a little pedantic , since all of these patterns can be subsumed under one rubric, viz:- disappearance. We are presently working on multi-stable perceptions where more than two alternatives are possible. These do not seem to present fatal difficulties to the hemispheric switching idea so far. In any case, multistable rivalries are relatively rare compared to the vast majority of rivalries, which are bistable.
Another apparent difference between binocular rivalry and Bonneh's MIB is that increasing the contrast seems to increase the duration of the dominant phase in opposite directions, since brighter yellow dots increase the disappearance phase while increasing contrast shortens the suppressed phase but has not effect on the dominant phase of binocular rivalry. We think that it is premature to make too much of this difference, since there are some forms of stimulus manipulation (e.g. changing the context) that CAN increase the dominance phase duration in binocular rivalry. The more monophasic form of the oscillation in Bonneh's MIB might align it more apporpriately with only one phase of the binocular rivalry oscillation, for example. In addition, brightening the yellow discs might be seen as a contextual stimulus in the sense that this would further separate the blue swirl from the discs and therefore tend to reject the hypothesis that the yellow discs were "connected" to the blue swirl.