Inhibition of return (IOR) refers to slower reaction times when a target appears unpredictably in the same location as a preceding cue, rather than in a different location. In the present study, frontal images of human faces were presented intact as face configurations, were rearranged to produce scrambled-face configurations, or were pixilated and randomized to produce nonface configurations. In an orienting paradigm designed to elicit IOR, face and scrambled-face stimuli were used as cues (Experiment 1), as targets (Experiment 2), or along with pixilated nonface stimuli as both cues and targets (Experiment 3). The magnitude of IOR for subsequent localization targets was unaffected by cue configuration. Likewise, the magnitude of IOR was unaffected by target configuration. These results suggest that IOR is a "blind" mechanism that is unaffected by the mere occurrence of biologically relevant cue and target stimuli.

A number of studies have presented evidence that middle spatial frequencies (SFs), between about 8 and 16 cycles per face width (c/fw), provide information that is more useful for face recognition than are other frequency ranges (Bachmann, 1991;Costen, Parker, & Craw, 1994Gold, Bennett, & Sekuler, 1999;Näsänen, 1999;Parker & Costen, 2001;Rolls, Baylis, & Hasselmo, 1987). On the basis of these data, it has been suggested that band-pass spatial filtering may be the first stage of feature extraction during visual recognition (Näsänen, 1999). That is, the visual information reaching purported face recognition areas might be filtered to preferentially represent middle SFs. In the extreme, it is possible that visual information reaching face recognition areas of the brain is first reduced to only middle SF ranges prior to further analysis. In theory, this filtering could help to minimize the processing load placed on face recognition mechanisms by discarding low and high SFs.A few recent studies have cast doubt on the idea of an absolute advantage for middle SFs in face recognition by showing that recognition performance is little affected by the range of frequencies contained in face stimuli, so long as both comparison and test stimuli are filtered in the same way (Collin, Liu, Troje, McMullen, & Chaudhuri, 2004;Kornowski & Petersik, 2003;Liu, Collin, Rainville, & Chaudhuri, 2000). These findings seem incompatible with the idea that middle SFs provide the most useful information for face recognition. If this were the case, one would expect some degree of performance loss when attempting to match two faces that contain no SFs from inside the middle band, relative to matching two faces with SFs from within the middle band. However, this difference has not been observed. Instead, Collin and colleagues (Collin et al., 2004) have suggested that a broad range of SFs may be useful for any visual task and that the range will depend on such task characteristics as similarity between learned and tested stimuli. This is a view echoed by Schyns and colleagues (Schyns, 1998;Schyns & Gosselin, 2003), who suggested that a wide range of SFs may be used in any given task and that there is a top-down control of which SFs are accessed in an image.In the following, we first will discuss two sets of studies that respectively provide evidence for and against the idea of a middle-band advantage in face recognition. We then will present a pair of experiments that examined how the differing task demands of these two sets of studies might lead to different conclusions about the role of SFs in face recognition.Rolls and colleagues (Rolls et al., 1987) provided some of the earliest evidence that face recognition might rely preferentially on middle SFs. They recorded activity of face-sensitive neurons in the macaque superior sulcus in response to band-pass filtered face images. They found that these neurons responded most strongly to images containing frequencies between 4 and 32 c/fw. Bachmann (1991) provided the first clue that these physiolog...

Previous studies have suggested that physiological responses are greatest and face recognition performance is best when a band of middle relative spatial frequencies (SFs) is included in stimuli. Conversely, behavioural data suggest that object recognition performance shows comparatively little effect of SF variations. Here, we examine the effects of SF filtering on the amplitude of the N170 ERP component when participants are shown images of faces and objects. Our findings show that with face stimuli the amplitude of N170 exhibits a band-pass modulation function, with responses to middle SFs (around 11 cycles per face) being statistically indistinguishable from responses to full-band faces. In contrast to faces, object stimuli elicited a relatively flat function across much of the spectrum. However, for both faces and objects, middle spatial frequencies were sufficient to elicit the same N170 magnitude as full-band images. Our results with face stimuli are in accordance with previous work examining single-cell and MEG responses. Our results with objects are compatible with previous behavioural work showing a relative robustness of object recognition to SF manipulations. Our findings are novel in showing that the middle band elicits the same N170 as full-band images in both faces and objects.

When a target appears unpredictably in the same rather than a different location relative to a preceding onset cue, reaction times (RTs) of participants tasked with responding to the target are slowed. This pattern of results, referred to as inhibition of return (IOR), is believed to reflect the operation of a mechanism that prevents perseverative search of nontarget locations. On the grounds that an evolved mechanism might be sensitive to social stimuli, Taylor and Therrien (2005) examined IOR for localization responses under conditions in which cues and targets could be intact face configurations or nonface configurations; contrary to their predictions, there was no influence of cue or target configuration on the magnitude of IOR, indicating that the mere occurrence of task-irrelevant face and nonface stimuli does not alter IOR. In the present study, we further examined this issue in a task that required a face/nonface target discrimination. When target configuration was thereby made task relevant, we found that IOR differed for face and nonface targets in terms of magnitude (when a single cue-target stimulus onset asynchrony was employed) and time course. We suggest that the RT delay associated with IOR may enable additional processing time and/or response selection when a task-relevant face is presented at the cued location.

No abstract