Figure 5.

Consistency of rat recognition strategy across object views. (A) Illustration of the procedure to compute the raw and aligned overlap between the salient features' patterns obtained for two different views of an object. The default and the leftward horizontally shifted views of Object 3 are used as examples (first row). The raw features' overlap was computed by superimposing the images of the two object views (and the corresponding features' patterns) within the stimulus display (second row, left plot). The aligned features' overlap was computed by reversing the transformation that produced the leftward horizontally shifted view. That is, the object was shifted to the right of 18° and scaled back to 35°, so as to perfectly overlap with its default view (second row, right plot). In both cases, the overlap was computed as the ratio between the orange area and the sum of the red, yellow and orange areas. The significance of the overlap was assessed by randomly shifting the salient regions of each object view within the minimum bounding box (see white frames in the third row of the figure) enclosing each view. (B) For each pair of views of Object 3 (circles) and Object 4 (diamonds) resulting from affine transformations (i.e., position/size changes and in-plane rotations), the raw features' overlap is plotted against the aligned features' overlap. The shade of gray indicates whether the raw and/or the aligned overlap for a given view was significantly larger than expected by chance (p < 0.05; see caption). (C) Median aligned overlaps for the objects belonging to Stimulus Set 1 and 2. The error bars are standard errors of the medians (obtained by bootstrapping). The statistical significance of the difference between a given pair of medians was assessed by a Mann–Whitney U-test (^****p < 0.0001). (D) Percentage of significant aligned overlap values for the objects belonging to two stimulus sets.