Representing invisible visual input
Not all information that is projected on the retinae of our eyes necessarily leads to conscious visual experience. Some information is perceptually suppressed for instance as the result of binocular conflict. Learning about the extent to which suppressed visual information is in fact still processed may contribute to our understanding of the functioning of visual awareness. In two experimental studies we have used different approaches to study how the processing of perceptually suppressed visual information can be affected.
Learning to suppress visual information
When incompatible information is presented to the left and the right eye, both images compete for visual awareness, leading to the perceptual suppression of one of the 2 images. Perceptual learning research has taught us that training a task on a certain stimulus generally sharpens its representation. In a similar fashion, we were interested in whether training to repetitively suppress one and the same stimulus will lead to an increase in suppression of this stimulus.
Experiment 1
Methods
Participants were trained for 4 sessions (over 2000 trials) on the suppression of an oblique grating that was always presented with the same orientation to the same eye. A high contrast expanding bull's eye prevented the visibility of the grating. Before and after training we measured contrast detection performance for gratings with the same or the orthogonal orientation as the suppression-trained grating, both for the trained as for the untrained eye. |
Results
On the right you see the threshold elevation for all four condition, where a high score reflects a relatively worse performance after training than before training. The results show that for both the trained eye as well as for the untrained eye threshold elevation is significantly higher for the trained orientation than for the untrained orientation. This effect indicate that indeed suppression of the trained orientation becomes relatively deeper as the result of suppression training. The fact that this effect also occurs for the untrained eye, suggests that this effect occurs beyond a monocular stage of visual processing. In addition threshold elevations are higher for the trained eye than for the untrained eye, reflecting an eye-specific suppression effect. |
Experiment 2
Methods
We performed a second experiment in which we applied a double-training protocol with the aim to investigate the contribution of monocular and/or a binocular processing stages to the effect of suppression learning. The procedure was similar to Experiment 1, but we doubled the number of training blocks. In half of the blocks participants were trained on suppressing one orientation in one eye, while in the other blocks they were trained on suppressing the orthogonal orientation in the other eye. Again we measured detection performance before and after training for four conditions. |
Results
As in Experiment 1 we computed threshold elevation for all conditions. When we collapse the data of both eyes, then we see that threshold elevation is significantly higher for the trained orientation than for the untrained orientation. This finding indicates that there is a monocular contribution to suppression learning, because if suppression learning would be purely binocular, then the effects of the double training would have cancelled each other out for the two trained orientations. Overall these experiments show that binocular suppression can be increased through training, specifically for the trained orientation. In addition, both monocular and binocular visual areas seem to be involved in this suppression learning. |
Published as Vergeer, M., Wagemans, J., & van Ee, R. (2016). Training of binocular rivalry suppression suggests stimulus-specific plasticity in monocular and binocular visual areas. Scientific Reports, 6, 25753.
Integrating visible and invisible stimulus parts into coherent visual representations
Perceptual integration has traditionally been investigated by having subjects perform tasks on the objects they consciously perceive. However, the exact relation between perceptual organisation and visual awareness has not yet been well established. For instance, it is unclear whether awareness of the different stimulus parts is required for visual integration to complete. We performed a study where we investigated whether invisible stimulus parts can be integrated with visible stimulus parts into one single coherent visual representation.
In the original version of a bistable, rotating cylinder, dots are moving horizontally along a vertical axis following a sinusoidal trajectory. This generally gives the impression of a rotating cylinder that can be perceived to rotate in a clockwise or in a counter-clockwise direction. At each point in time, half of the dots are perceived as being part of the convex front surface, while the other half are perceived as being part of the concave back surface. We split up this stimulus by presenting the leftwards moving dots in the right eye and the rightwards moving dots in the left eye. This caused binocular rivalry, in which only the dots presented to one eye were visible, while the dots presented to the other eye were perceptually suppressed.
|
Methods
Participants viewed the above described stimulus for trials of 120 seconds, and were asked to track their percept with keyboard presses. We were particularly interested in perceptual switches from leftward to rightward moving surfaces, and vice versa. We hypothesised that the invisible stimulus parts integrated with the visible stimulus halves into one coherent rotational representation. Following this hypothesis, we predicted that a perceptual eye switch would also lead to a curvature switch (from convex to concave, or vice versa), as a convex representation of one cylinder half goes hand in hand with a concave representation of the other (in this case suppressed) cylinder half. |
Results
We computed the percentage of eye switches in which the visible representation before and after the switch were consistent with the same visual representation. This percentage was computed for the already described binocular rivalry condition, as well as for a control condition in which only half of the stimulus was presented at a time. For almost all subjects this percentage was much higher in the rivalry condition, which supports our hypothesis that the suppressed half of the cylinder, although not visible to the observer, is represented by the visual system in a state that is consistent with the state of the visible half of the cylinder. In other words, complex visual integration processes do not require conscious access to all visual input to which they are applied to. |
Published as Vergeer, M., Moors, P., Wagemans, J., & van Ee, R. (2016). Visible and invisible stimulus parts integrate into global object representations as revealed by combining monocular and binocular rivalry. Journal of Vision, 16(11):15, 1-10.