Fig 1.
Graphic representation of a trial.
Each trial started with the presentation of three objects, of which the FT (here: triangle) was foveated. After an auditory go cue, a horizontal saccade was initiated to the ST (here: circle). Upon detection of the saccade, two objects were removed from the screen while the other was displaced (here: orthogonal). The displaced object remained visible for 50 ms, 300 ms or until the response was given (~1000 ms). The remembered presaccadic location of this object was indicated using a computer mouse.
Fig 2.
Performance of a single participant.
Red dots represent localization responses and blue shaded areas represent the response probabilities, p(s|mv), according to the best-fit model predictions. Left three columns, localization errors for parallel target displacements; right three columns, errors for orthogonal displacements. Horizontal dashed line represents veridical localization, i.e. the segregation strategy. Dashed diagonal line represents the displacement of the postsaccadic target. With small displacements, errors deviate toward the diagonal line for the three targets; this pulling effect appears stronger with longer viewing durations.
Fig 3.
Mean localization errors across participants.
Mean responses are shown as dots (error bars, SEM) and mean model fits as continuous lines (shaded areas, SEM). Format as in Fig 2.
Table 1.
Best-fit parameter values for all eleven participants.
All values are in degrees except probability PC. Position of π is expressed relative to FT.
Fig 4.
(A) Average σm across participants (error bars, SEM). The orthogonal component of the memorized positions appears to be more precise than the parallel component for FT and ST, but not for NT. (B) Average σvacross participants. Variability of the postsaccadic-target representation decreases as a function of viewing duration. (C) Prior π, positioned relative to FT, representing where objects are generally expected to appear. All values are in degrees.
Fig 5.
Inferred probability of a common cause p(C|mv) as a function of target displacement.
Probabilities are based on the best-fit parameters, separated by target location (rows: FT, ST, NT), displacement directions (columns: parallel/orthogonal) and postsaccadic viewing duration(in color). Shown are the mean values across participants and standard error (shaded areas). This probability, which can be interpreted as the complementary probability of perceiving the displacement, optimally weights the integration and segregation strategy.
Fig 6.
The presaccadic location of the NT (square) is reported after a transsaccadic displacement of 10° to the right. Objects in red represent visible targets; the white objects depict the veridical target locations. Representations of location estimates, modeled as 2D Gaussians, are shown as dark ellipses. (A) Before the saccade, all three objects are encoded with the foveal prior f (light grey blob) being centered at the triangle, the FT. After the saccade, the displaced target’s position and identity (NT here) are encoded with f now being centered at the saccade landing position. (B) Based on the NT’s presaccadic (m) and postsaccadic (v) representations, both biased by f, the probability of a single stable object, p(C|mv), is computed. In case m and v are unrelated the best solution is to segregate and ignore v. If m and v derive from the same object, the best solution is to integration all signals. (D) The two solutions in (C) are weighted according to the probability that m and v are related. The localization response follows from .