Fig. 4.

Figure–ground discrimination STAFs constructed for a stimulus in which a randomly striped figure moved against a randomly striped ground on separate m-sequences. (A) For compound figure–ground stimuli, wing optomotor steering responses correlated with the movement of the figure. Impulse responses are parameterized to the azimuthal position of the bar (e.g. responses appearing on the midline are calculated when the bar is directly in front of the fly). (B) The figure response measured with the figure in the front (black) and rear (gray) of the arena, indicated by the arrowheads in A. Inset: figure responses compared with and without concomitant ground motion [red line; re-plotted from Aptekar et al. (Aptekar et al., 2012)]. (C) For compound figure–ground stimuli, wing optomotor steering responses correlated with the motion of the wide-field ground parametized by the azimuthal position of the bar. (D) The ground response measured with the figure in the front (black) and rear (gray) of the arena from the positions indicated by the arrowheads in C. Inset: ground responses re-plotted from D (gray) for the random striped pattern compared with a yaw pattern that contains no figure [red; re-plotted from Theobald et al. (Theobald et al., 2010a) on the same time scale]. (E) Azimuthal profile of the STAFs summed over the first 100 ms, highlighting the spatial variation of the impulse response amplitudes. Black line, figure; gray line, wide-field ground; dashed black line, mean of figure and ground. Inset: figure STAF spatial profiles compared for concomitant ground motion (E) and a stationary ground [red line; re-plotted from Aptekar et al. (Aptekar et al., 2012)]. N=27 flies.