Contextual effects of scene on the visual perception of object orientation in depth

Abstract

We investigated the effect of background scene on the human visual perception of depth orientation (i.e., azimuth angle) of three-dimensional common objects. Participants evaluated the depth orientation of objects. The objects were surrounded by scenes with an apparent axis of the global reference frame, such as a sidewalk scene. When a scene axis was slightly misaligned with the gaze line, object orientation perception was biased, as if the gaze line had been assimilated into the scene axis (Experiment 1). When the scene axis was slightly misaligned with the object, evaluated object orientation was biased, as if it had been assimilated into the scene axis (Experiment 2). This assimilation may be due to confusion between the orientation of the scene and object axes (Experiment 3). Thus, the global reference frame may influence object orientation perception when its orientation is similar to that of the gaze-line or object.

Figure 1. Examples of the stimuli.

A. Scene Absent condition and Scene 0° condition in Experiment 1. B. Six scenes shown in gaze-aligned orientations. All stimulus images were presented in color during the experiment.

Figure 2. Schematic illustration of the apparatus used in Experiments 1 and 2.

A. Participants rotated the disk on the horizontal response display so that the disk orientation matched the perceived depth orientation of the object. B. On the response display, a white dot indicated the frontal orientation of the response disk. The disk orientation θ was measured as evaluated orientation. The characters, arrow, and dotted lines did not appear in the experiments.

Figure 3. Frequency distributions of evaluated orientation in Scene 0° condition, Experiment 1.

Bin width is 6°. The six distributions correspond to the six object orientation conditions, as indicated by the symbols. The vertical dotted lines mark the positions of stimulus object orientation (the “true” orientation). The vertical solid lines mark the mean evaluated orientations. The evaluated orientations were biased toward profile, as indicated by the arrows. Since the results were symmetric, we ignored left/right difference and symmetric object orientation conditions were merged into single condition in the following analyses. L, left; R, right.

Figure 4. Results of the object orientation evaluation task in Experiment 1.

Scene orientation was manipulated relative to the gaze line. Bias (evaluated orientation minus true object orientation) is plotted as a function of object orientation (A–C). The results of the Scene 0° condition are plotted repeatedly for comparison. Overall, a positive bias was observed, confirming the bias toward profile in the perception of oblique object orientations. Importantly, the bias was modulated by scene condition. Panels D and E illustrate top views of the spatial layouts of objects and scenes in the conditions indicated by the arrows in panel B. Dotted lines with open triangles indicate subjective gaze lines that might be biased toward the scene orientations. Panels F–G indicate the SD of the evaluated orientation as a function of object orientation. The ANOVA (object orientation × scene condition) was conducted for each of the panels A–C and F–G. The significant main effect and simple main effect of scene condition are marked by asterisks. **, p<.01, *, p<.05. Scn. = scene; obj. = object.

Figure 5. Results of the object orientation evaluation task in Experiment 2.

Scene orientation was manipulated relative to the object. The bias (evaluated orientation minus object orientation) is plotted as a function of object orientation (A–C). The results from the Scene 0° condition are plotted repeatedly for comparison. The bias toward profile was confirmed just as in Experiment 1. Further, +/−9° scenes modulated the magnitude of this bias (B). The spatial layouts of the conditions in which +/−9° scenes affected the results are schematized in panels D and E. As indicated by the dotted arrows, object orientations were evaluated as if they were assimilated into the scene orientations. Panels F–G indicate the SD of the evaluated orientation as a function of object orientation. The ANOVA (object orientation × scene condition) was conducted for each of the panels A–C and F–G. The significant main effect and simple main effect of scene condition are marked by asterisks. **, p<.01, *, p<.05. Scn. = scene; obj. = object.

Figure 6. Results of Experiment 3.

Participants were asked to report the direction of misalignment between the object and scene axes. Scene orientation was misaligned +/−9° or +/−18° with respect to object orientation. Scene-object misalignments were more easily perceived for objects oriented at 9° than for objects oriented at 45°.