Abstract
Humans can reliably recognize faces across viewpoints. Although suboptimal, their performance is quite remarkable given the large changes in low-level image properties a shift in viewpoint introduces. We used a behavioral and an fMRI adaptation experiment to investigate whether this viewpoint tolerance is reflected in the neural visual system and whether it can be manipulated through training. Participants saw training sequences of faces displayed in 7 different rotation angles, creating the appearance of a rotating head. Half of the sequences showed faces undergoing veridical changes in appearance across the rotation (non-morph condition). The other half were non-veridical: during rotation, the face simultaneously morphed into a second face. This procedure should succesfully link frontal face views to side views of the same or a different identity, and, according to the temporal contiguity hypothesis, thus enhance viewpoint-invariance in the non-morph condition and/or break invariance in the morph condition.Test phase trials contained two sequential face images – a frontal and a side view. The fMRI experiment also contained trials with untrained faces and same-viewpoint trials. Performance on the same/different task in the behavioral experiment (N=20) was affected by training. There was a significant interaction (p=0.03) between training (linked/not linked) and identity (same/different), reflecting a selective drop in performance for different identity trials of morphed faces. In the fMRI study (N=20), fMRI adaptation effects were found for same-viewpoint images of untrained faces, but no adaptation for untrained faces was present across viewpoints. Only faces which were not morphed during training elicited a slight adaptation across viewpoints in face-selective regions (FFA: p=0.045, OFA: p=0.03, one-tailed t-test). No such effect was present for morphed faces (p=0.55 and p>0.99). Behavioral and neural results are therefore compatible: Temporal contiguity can influence viewpoint tolerance, with more evidence for tolerance when faces are not morphed during training.
Meeting abstract presented at VSS 2017