Visual Search of Mooney Faces

Abstract

Faces spontaneously capture attention. However, which special attributes of a face underlie this effect is unclear. To address this question, we investigate how gist information, specific visual properties and differing amounts of experience with faces affect the time required to detect a face. Three visual search experiments were conducted investigating the rapidness of human observers to detect Mooney face images. Mooney images are two-toned, ambiguous images. They were used in order to have stimuli that maintain gist information but limit low-level image properties. Results from the experiments show: (1) Although upright Mooney faces were searched inefficiently, they were detected more rapidly than inverted Mooney face targets, demonstrating the important role of gist information in guiding attention toward a face. (2) Several specific Mooney face identities were searched efficiently while others were not, suggesting the involvement of specific visual properties in face detection. (3) By providing participants with unambiguous gray-scale versions of the Mooney face targets prior to the visual search task, the targets were detected significantly more efficiently, suggesting that prior experience with Mooney faces improves the ability to extract gist information for rapid face detection. However, a week of training with Mooney face categorization did not lead to even more efficient visual search of Mooney face targets. In summary, these results reveal that specific local image properties cannot account for how faces capture attention. On the other hand, gist information alone cannot account for how faces capture attention either. Prior experience facilitates the effect of gist on visual search of faces; making faces a special object category for guiding attention.

Stimulus-Specific Individual Differences in Holistic Perception of Mooney Faces

“…The stimuli consisted of 192 images (96 ambiguous Mooney faces and 96 shuffled Mooney stimuli). Literally thousands of Mooney faces have been created and published in the past (Brodski et al, 2015;Goold and Meng, 2016;Ke et al, 2017;Schwiedrzik et al, 2018). In the present study, we used a subset of Mooney faces created by Schwiedrzik et al (2018) because these Mooney faces have been used to study individual differences, and because these stimuli include a diverse array of faces (varying in perspective, lighting direction, face identity, gender, etc.).…”

“…There are a large variety of Mooney faces that have been created and published, including the original artist-created images (Mooney, 1957), thresholded images of gray-scale faces (Brodski et al, 2015;Goold and Meng, 2016;Schwiedrzik et al, 2018), and artificially created images based on machine learning approaches (Ke et al, 2017). While, Mooney faces are typically treated as an excellent stimulus to isolate holistic processing, there are also significant variations between different Mooney faces (Figure 1).…”

How prior knowledge prepares perception: Alpha-band oscillations carry perceptual expectations and influence early visual responses

“…These behavior results are novel in two respects. First, it marks the first demonstrations, to our knowledge, of cuing recognition of Mooney-style images using solely linguistic cues, as opposed to the more common method of simply revealing the original image 17,18,52 . Second, the results of our same/different discrimination task reveal that linguistic cues enhance not only the ability to recognize the images, as in prior work, but also putatively lower-level processes subserving visual discrimination.…”

“…The two conditions differed only in how participants were familiarized with the images. In the meaning trained condition, participants first viewed each Mooney image accompanied by an instruction, e.g., "Please look for CAKE", twice for each Mooney image (Trials [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20]. Participants then saw all the images again and were asked to type in what they saw in each image, guessing in the case that they could not see anything (Trials 21-30).…”

“…There is now considerable evidence, however, suggesting that prior knowledge impacts relatively early stages of perception [6][7][8][9][10][11][12][13][14][15] . A dramatic demonstration of how prior knowledge can create meaning from apparently meaningless inputs occurs with two-tone "Mooney" images 16 , which can become recognizable following the presentation of perceptual hints 17,18 .…”