Recent experimental results characterizing the face processing network in macaque visual cortex pose a major puzzle. View-tuned units (found in patches ML/MF) are a natural step to a view-tolerant representation (found in patch AM), as predicted by several models. However, the observation that cells in patch AL are tuned to faces and their mirror reflections remains unexplained (cf. Freiwald and Tsao, 2010 & Leibo et al. 2011). We show that a model based on the hypothesis that the ventral stream implements a memory-based approach to transformation invariance predicts the main properties of ML/MF, AL and AM.
View-invariance and mirror-symmetric tuning in a model of the macaque face-processing system
Fabio Anselmi;
2015-01-01
Abstract
Recent experimental results characterizing the face processing network in macaque visual cortex pose a major puzzle. View-tuned units (found in patches ML/MF) are a natural step to a view-tolerant representation (found in patch AM), as predicted by several models. However, the observation that cells in patch AL are tuned to faces and their mirror reflections remains unexplained (cf. Freiwald and Tsao, 2010 & Leibo et al. 2011). We show that a model based on the hypothesis that the ventral stream implements a memory-based approach to transformation invariance predicts the main properties of ML/MF, AL and AM.File in questo prodotto:
Non ci sono file associati a questo prodotto.
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
