Purchase this article with an account.
Fang Fang, Junzhu Su, Cheng Chen, Dongjun He; Perceptual learning induces fast processing and efficient representation of face. Journal of Vision 2011;11(11):589. https://doi.org/10.1167/11.11.589.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Perceptual learning can significantly improve the capability of the adult visual system for object recognition and discrimination, yet little is known about how learning affects object processing and representation in the brain. Here, we combined EEG and psychophysics to study the neural mechanism of perceptual learning of face view discrimination. We trained subjects to discriminate face in-depth orientations at a face view (i.e. 30°) over eight daily sessions (8000 trials in total), which resulted in a significant improvement in sensitivity to the face view orientation. Before and after the 8-day training, we measured subjects' face orientation discrimination thresholds at the face views of −90, −60, −30, 0, 30, 60, and 90° (minus means left tilt), and EEG signals induced by the trained and untrained face views were recorded. We found that this improved sensitivity was highly specific to the trained view. EEG data showed that, after training, the peak latency of N170 at the left occipito-temporal area evoked by the trained view was significantly shortened (about 8 ms) and the 50–90 Hz gamma power at the left prefrontal area induced by the trained view was significantly reduced between 100 and 200 ms after stimulus onset. Parallel to the psychophysical finding, the N170 latency shortening and the gamma power reduction were also specific to the trained face view. Previous researches have demonstrated the close relationship between N170 and face configural perception and the existence of face selective neurons/areas in the prefrontal area. Some theoretical and empirical studies suggest that gamma power reduction means a more efficient cortical representation of a stimulus. We conclude that face view discrimination training could lead to faster processing and a more efficient representation of face. These results provide important constraints on the neural model of high-level visual perceptual learning.
This PDF is available to Subscribers Only