Figure 1.

Performance on a 3D motion perception task, and the impact of active viewing. (a) Experimental procedure without feedback. Observers judged the 3D direction of a moving dot (target). The target appeared in the center of the display at the start of a trial. After the target traversed a randomly chosen trajectory (‘Presented direction’), it disappeared and a response ‘paddle’ appeared. Observers adjusted the position of the paddle around a circular orbit so that it would have intercepted the target had it continued along its trajectory (‘Reported direction’). No feedback was provided. (b) Illustration of left- and right-eye stimulus elements as presented in the experiments. (c) Reported motion direction as a function of presented direction for all 38 observers (75 trials each, targets presented at mid contrast under the fixed viewing condition). Data points along the positive diagonal correspond to accurate reports of 3D motion trajectory. Deviations from that diagonal correspond to misreports of motion trajectories. (d) Performance (% of target interceptions) as a function of trial number. An interception was defined as a reported target direction that would hit the paddle (within +/− 8 deg of the presented direction). % Interceptions did not differ significantly between the fixed viewing condition (blue symbols) and active viewing (motion parallax) condition (green symbols), either at the beginning of the experiment, or with repeated exposure. Data points correspond to the percentage of target interceptions across observers on each trial; solid lines correspond to the exponential fit to the data, and the shaded regions correspond to the 95% confidence intervals on the exponential fit.