Adaptation Effects from Inverted Filtered Faces: Testing Specificity Across Identity Levels

Description This follow-up study investigates whether the adaptation effects observed in our first blurring and sharpening experiment are specific and sensitive to faces or whether they may result from more general, image-based visual processing. To test this, we employed a face adaptation paradigm using inverted (upside-down) facial images as adaptation stimuli. It is well established that face inversion significantly disrupts face recognition while largely preserving low-level visual features such as color, shape, and contrast. Thus, inverted faces serve as a meaningful control condition to assess the specificity of face-related perceptual mechanisms. As in our first study, participants are first exposed to heavily manipulated images of celebrity faces (using blur filters or sharpen filters). However, in this version of the experiment, the adaptation stimuli – not the test stimuli – are presented in an inverted orientation. During the test phase, participants are asked to choose between an original image and a subtly altered version. An adaptation effect would be indicated by a preference for the slightly modified image (modified in the direction of the adaptation stimulus), suggesting that the perceptual system had adapted to the previously viewed extreme version. Importantly, this experiment used inverted versions of the same adaptation images that were used in the first study with upright faces. This allows for a direct comparison between the two experiments – specifically between potential face adaptation effects under upright versus inverted viewing conditions. By using the same adaptation images as in the first study (but now shown in inverted orientation), we minimize low-level visual confounds and can better assess whether observed effects truly rely on face processing or can also occur when such processing is disrupted. As before, adaptation effects were measured across three conditions: (1) pictorial (same image, same identity), (2) structural (different image, same identity), (3) cross-identity (different image, different identity). The results of this study aim to clarify whether the previously observed adaptation effects are driven by high-level mechanisms of face perception or whether they are attributable to more general visual adaptation processes.