Neural basis of familiar face recognition

Familiar identities can be easily recognised across different images. This contrasts with the difficulty in recognising unfamiliar identities across changes in image. Thus, cognitive models propose different neural processing of familiar and unfamiliar faces: familiar identities are suggested to be represented independent of face image (image-invariantly) while unfamiliar identities are suggested to be represented more image-dependently. This image-invariance proposed for familiar faces is suggested to underlie the familiar face recognition advantage. However, the neural correlate of this proposed difference in neural representation of familiar and unfamiliar identities remains unclear. This thesis presents multiple fMRI experiments conducted to investigate the neural correlate of familiar face recognition. The first empirical chapter shows a link between activity in the fusiform face area (FFA) and the familiar face recognition advantage. The second and third empirical chapters investigate the representation of faces within the FFA as well as other face-selective regions. The last empirical chapter demonstrates that the image-invariant representation proposed to underlie the familiar face recognition advantage seems to lie outside the FFA and other core face-selective regions. Taken together, these findings support the suggested role of the FFA in identity processing. Additionally, we show that regions outside the core face-selective regions might be necessary for the recognition of familiar identities.