The modulatory effects of facial cues and familiarity in face recognition: a behavioral and eye-tracking investigation.

Introduction: Face recognition is essential for social interaction, yet the extent to which internal facial features and external facial cues contribute to identity recognition remains incompletely understood. In particular, it remains unclear how external appearance changes, such as hairstyle and hair color, influence the recognition of familiar and unfamiliar faces across perceptual and memory-based processing stages. Methods: This study examined the modulatory effects of external facial cues and face familiarity using behavioral measures and eye-tracking. Thirty-four native Chinese-speaking adults completed a modified visual paired comparison face memory task, and 31 of them also completed a face matching task. Facial stimuli included familiar Chinese celebrities and unfamiliar faces, with systematic manipulations of hairstyle and hair color. In the face matching task, participants judged whether two simultaneously presented faces depicted the same identity. In the face memory task, participants first encoded a single face and then judged recognition when the familiarized face was paired with either a different identity or the same identity with altered external cues. Eye movements were analyzed across predefined areas of interest, including internal features, internal non-feature regions, and external regions. Results: In the face matching task, external cue variations did not significantly affect accuracy but increased reaction times, particularly for unfamiliar faces. Reaction times were longer when hairstyle or hair color differed, and hair color changes showed a stronger disruptive effect for unfamiliar than familiar faces. In the face memory task, hairstyle manipulations selectively affected recognition accuracy for unfamiliar faces, whereas hair color manipulations did not show the same pattern. Eye-tracking results showed that participants consistently prioritized internal facial features across tasks. However, gaze allocation and gaze transition patterns were flexibly modulated by face familiarity and external cue variations. During memory encoding, familiar faces elicited relatively greater attention and transitions involving external regions, whereas unfamiliar faces were associated with a more internally focused scanning pattern. During recognition, external cue changes, especially hairstyle changes, increased dwell time on the target face and altered comparison-related gaze dynamics. Discussion: These findings suggest that face recognition is a dynamic process shaped by the interaction between external facial cues, familiarity, and task demands. Internal facial features remained the primary source of identity-relevant information, but external cues influenced processing efficiency, memory-based recognition, and visual exploration. Familiar faces appeared more robust to external appearance changes, whereas unfamiliar faces were more susceptible to disruptions caused by changes in external cues, particularly hairstyle. The results provide a more comprehensive account of how facial identity information is extracted and integrated across perceptual and memory-based recognition stages, and offer a foundation for future research on face recognition in populations with atypical face processing.