Faces in the Mist: Illusory Face and Letter Detection

Abstract

We report three behavioral experiments on the spatial characteristics evoking illusory face and letter detection. False detections made to pure noise images were analyzed using a modified reverse correlation method in which hundreds of observers rated a modest number of noise images (480) during a single session. This method was originally developed for brain imaging research, and has been used in a number of fMRI publications, but this is the first report of the behavioral classification images. In Experiment 1 illusory face detection occurred in response to scattered dark patches throughout the images, with a bias to the left visual field. This occurred despite the use of a fixation cross and expectations that faces would be centered. In contrast, illusory letter detection (Experiment 2) occurred in response to centrally positioned dark patches. Experiment 3 included an oval in all displays to spatially constrain illusory face detection. With the addition of this oval the classification image revealed an eyes/nose/mouth pattern. These results suggest that face detection is triggered by a minimal face-like pattern even when these features are not centered in visual focus.

Seeing Jesus in toast: Neural and behavioral correlates of face pareidolia

“…Behavioral reverse correlations showed that the face CI produced a face-like structure, consistent with previous reverse correlation studies (Gosselin & Schyns, 2003; Hansen et al, 2010; Rieth et al, 2011), even though we used relatively fewer trials than those existing studies. More importantly, the spectral properties of our face CI significantly correlated with those of a noise-masked averaged face image but did not correlate with the letter CI or the noise-masked averaged letter image.…”

“…In the present study, we use the noise images which were produced by overlapping the random-positioned bivariate Gaussian blobs with SDs of 64, 256, and 1024 in order to reliably generate the face pareidolia and letter pareidolia, respectively. The reason that we used Gaussian blobs with multiple SDs to create complex random noise images was that existing behavioral studies using this method reliably induced face or letter paradolia at about 35% of the time (Rieth et al, 2011; Liu et al, 2010). Pilot testing also showed that illusory detections of faces or letters substantially decreased when simple random noise images were used, resulting in too few illusory trials for meaningful data analyses.…”

“…Recent behavioral and functional imaging studies using a reverse correlation method have provided some intriguing insights about how face pareidolia might emerge. These studies have demonstrated that the internal representation of faces underlying face pareidolia can be reconstructed experimentally based on behavioral responses (Gosselin & Schyns, 2003; Rieth, Lee, Lui, Tian & Huber, 2011; Smith, Gosselin, & Schyns, 2012) or on brain activities measured by EEG (Hansen, Thompson, Hess, & Ellemberg, 2010) and fMRI (Nestor, Vettel, & Tarr, 2012). For example, in Gosselin and Schyns’ (2003) study, participants were instructed to detect a face from pure-noise images where, in fact, no face images existed.…”

Stimulus Requirements for Face Perception: An Analysis Based on “Totem Poles”

“…(It is possible that in addition to symmetry there was a bias to look for faces along the vertical midline or “middle” of the image. However, Figures 7 and 9 suggest this bias must be a much weaker factor, and moreover as noted above, Rieth et al (2011) instead found a left field bias in asymmetric noise). Symmetry is a highly conspicuous feature that produces strong activity in higher visual areas (Sasaki et al, 2005), and is likely to be an important configural dimension in face coding (Rhodes et al, 2005).…”

“…For example, Gosselin and Schyns (2003) assessed the perception of smiles in mouthless faces with added noise. Zhang et al (2008) explored patterns of BOLD activation when observers looked for faces in pure noise, and recently extended this approach to examine behavioral responses to illusory faces (Rieth et al, 2011). They found that when an oval context for a head outline was provided, face percepts were triggered by simple configurations consistent with prominent internal (eyes, nose, mouth) but not external (e.g., ears) features, similar to the basic template we inferred.…”