Abstract
Adaptation to simple shapes and forms can produce shape-contrast after-effects (Suzuki and Cavanagh, 1998). Adaptation to complex visual stimuli such as faces can likewise produce contrastive figural after-effects (Webster and MacLin, 1999). Here, we examine the transfer of adaptation from simple visual forms to human face stimuli (Exps. 1 and 2) and vice-versa (Exp. 3).
In Experiment 1, observers adapted to a vertical ellipse, a horizontal ellipse, or a perfect circle. Normal faces were judged to be horizontally elongated following adaptation to a vertical ellipse and vertically elongated following adaptation to a horizontal ellipse. Adaptation to the circle did not distort the perception of faces.
In Experiment 2, observers adapted to pairs of convex and concave arcs, e.g., ( ) and ) ( respectively. We varied the aspect ratio of the arcs pairs, changing the height of the arcs and the width between the arcs, so that they appeared vertically or horizontally elongated. Following adaptation to narrow convex and concave arcs, observers judged faces as wider. An opposite after-effect occurred with widely spaced arcs. When observers adapted to a convex arcs with the average aspect ratio of normal faces the subsequent face perception remained unaltered. Combined, these results are consistent with Experiment 1 and further confirm that adaptation to simple visual forms can alter perception of faces.
In Experiment 3, observers adapted to faces that varied in their aspect ratio and were tested with circles. We found little evidence that face adaptation can alter the perception of the circle.
The present study suggests an asymmetric relationship in the transfer of adaptation aftereffects between high and low level visual stimuli. These findings are partially consistent with O'Leary and McMahon (1991) and lay a foundation for using adaptation to explore the organization of neural codes at different levels of visual abstraction.