Fig 1.
From left to right: hat, no hat, hatband, vest, target-absent; note the matched target size within each set, while there is variability between locations; for both examples, the target-absent scene matching the vest scene is depicted (these are distinct photographs, no image manipulation besides cropping was used in any case throughout this study). All persons depicted agreed in writing to the unrestricted use of their image, in particular to inclusion in the online database and to the use in psychophysical experiments. All images have been made available by the authors under a CC-BY-4.0 license.
Fig 2.
Creation of stimuli from database images.
A) Images from the database (original resolution 4912 × 2760 pixels) for the same model (#2 in the database); target-present images are scaled down such that the target size is identical across conditions using the smallest bounding box for the model as reference; the reference image (here the vest condition, second panel from bottom) remains unscaled; black frames are to ease size comparison and not part of the images. B) The eight target-present stimuli created from the example image on top (also forth image of panel A). The stimuli on the right are the mirror reflected versions of the stimuli on the left. The target-absent stimuli are created by randomly picking one of the cropping areas and applying it to the target-absent image taken at the same location as the model’s images (bottom of panel A), resulting in a total of 8 target-absent images per location; hence, there are a total of eight images of each condition per location (model). C) Histogram of the locations of the models’ bounding boxes (i.e., search targets) relative to the stimulus (and screen) for all 640 (20 models×4 conditions×8 versions) target-present stimuli. The person depicted in panels A and B agreed in writing to the unrestricted use of their image. All images have been made available by the authors under a CC-BY-4.0 license.
Fig 3.
A) Error rate by condition (only target-present conditions shown, target absent at 1.8%); B) Error rate for vest condition, split by vest color; C) Behavioral reaction times (RTs) by condition, target-absent data at 1.01s not depicted; D) RTs for vest condition, split by vest color. All bars denote mean and standard error of the mean (s.e.m.) across individuals. Significance markers in all panels refer to pair-wise comparisons (*p < 0.05; **p < 0.01; ***p < 0.001).
Fig 4.
A) Time to the first fixation on the person (eyeRT) by condition. B) Example for gaze data for two stimuli. Each colored line represents the gaze position data of a different individual, green and white filled circles fixations (green-first fixation, starting before stimulus onset). Top: hatband, bottom: vest. Note the frequent second saccades after fixating on the vest. C) eyeRT in the vest condition split by vest color. Bars denote mean and s.e.m. across individuals. Significance markers in all panels refer to pair-wise comparisons (*p < 0.05; **p < 0.01; ***p < 0.001). The persons depicted in panel B agreed in writing to the unrestricted use of their image. All images have been made available by the authors under a CC-BY-4.0 license.