Simulated forward and backward self motion, based on realistic parameters, causes motion induced blindness

Abstract

Motion Induced Blindness (MIB) is a well-established visual phenomenon whereby highly salient targets disappear when viewed against a moving background mask. No research has yet explored whether contracting and expanding optic flow can also trigger target disappearance. We explored MIB using mask speeds corresponding to driving at 35, 50, 65 and 80 km/h in simulated forward (expansion) and backward (contraction) motion as well as 2-D radial movement, random, and static mask motion types. Participants (n = 18) viewed MIB targets against masks with different movement types, speed, and target locations. To understand the relationship between saccades, pupil response and perceptual disappearance, we ran two additional eye-tracking experiments (n = 19). Target disappearance increased significantly with faster mask speeds and upper visual field target presentation. Simulated optic flow and 2-D radial movement caused comparable disappearance, and all moving masks caused significantly more disappearance than a static mask. Saccades could not entirely account for differences between conditions, suggesting that self-motion optic flow does cause MIB in an artificial setting. Pupil analyses implied that MIB disappearance induced by optic flow is not subjectively salient, potentially explaining why MIB is not noticed during driving. Potential implications of MIB for driving safety and Head-Up-Display (HUD) technologies are discussed.

Figure 1

Experimental display for Experiments 1, 2 and 3. Schematic black arrows were not shown on the display. They depict expanding motion shown on the left, and contracting motion shown on the right.

Figure 2

Experiment 1. Percentage, Mean Rate and Mean Duration of target disappearance dependent on movement type. Double asterisk (**) indicates significance at the 0.01 level; error bars indicate standard error (+/−1 SE).

Figure 3

Experiment 2. Percentage, Mean Rate and Mean Duration of target disappearance dependent on movement type. Double asterisk (**) indicates significance at the 0.01 level; error bars indicate standard error (+/−1 SE).

Figure 4

Experiment 2. Relationship between mean number of macro and macro+micro saccades and percentage of disappearance per trial for each mask condition.

Figure 5

Experiment 3. Percentage, Mean Rate and Mean Duration of target disappearance per trial for each mask speed. Single asterisk (*) indicates significance at the 0.05 level. Error bars indicate standard error (+/−1 SE).

Figure 6

Experiment 4. Percentage, Mean Rate and Mean Duration of target disappearance for each 2D mask movement condition. Double asterisk (**) indicates significance at the 0.01 level. Error bars indicate standard error (+/−1 SE).

Figure 7

Experiment 2 (A) and 4 (B). Phasic pupil diameter changes (in % modulation) in response to a reported disappearance (left) and reappearance (right). Different conditions are displayed with different colors. Significance is indicated below the plot (Green = contraction is significantly different from static, Blue = expansion is significantly different from static at 0.05 level. There was no significant difference between contraction and expansion). Shaded areas represent 1 SE.

Figure 8

Experiments 2 and 4. Mean tonic (baseline) pupil diameter (+/−1 SE) per mask condition.

Figure 9

Mean percentage of invisibility per trial according to target quadrant across all experiments. Similar trends were obtained for analyses using mean rate of target disappearance per trial.