Abstract
Introduction: When identifying a face, a majority of humans direct their initial eye movements to a featureless point just below the eyes. Such preferred point of fixation maximizes face identification accuracy and is predicted by a theoretical model that takes into account the distribution of information across the facial features and the foveated nature of the visual system (foveated ideal observer, Peterson & Eckstein, 2012). However, the visual attributes of the face that are processed in the visual periphery and guide the destination of the first fixation into faces are not known. Here, we manipulated the position of facial features and the head frame to assess which attributes are utilized by the brain to guide the 1st eye movement. Methods: Observers identified a face (14 deg. height) presented for 1 sec. out of 4 possible faces. Their starting fixation was 5-7 deg. from the edge of the face. Across conditions, the configuration of the features within the face was manipulated (e.g., top to bottom: Mouth, Nose, Eyes; etc.), the head frame was eliminated, or a single feature was erased. Observers were free to make eye movements. Results: The endpoint of the initial fixation was not altered when the mouth or nose was erased but was directed to lower point along the face when the eyes were eliminated. Similarly, lowering the position of the eyes within the face also lowered the destination of the 1st saccade. Finally, eliminating the head frame also lowered 1st saccade locations along the face. Conclusion: The results suggest that the human brain relies heavily on the position of the eyes and the head frame to guide the first saccade into faces. Future research should investigate how the features guiding eye movements might interact with the initial position of the face in the visual periphery.
Meeting abstract presented at VSS 2018