< Back to previous page

Project

The spatial and temporal structure of visual object space in the human brain

Visual object recognition is a crucial ability in many aspects of human interactions with the surrounding environment. Objects are represented in the lateral and ventral occipitotemporal cortex but it has proven to be difficult to distill an integrative view of the complex functional architecture of these regions. Bao et al. (Nature, 2020) proposed a comprehensive map of object space in inferior temporal cortex of monkeys: the inferior temporal cortex is organized as a map with two main dimensions, stubby-spiky and animate-inanimate. However, the study suffers from some limitations that we aim to resolve. it is unclear to what extent we can extrapolate from monkeys to humans. While overall their stimulus set dissociated stubby-spiky from animate-inanimate, this was not true within stimulus classes and the other visual dimensions were not controlled between animate and inanimate stimuli. Moreover, they considered a unique object space along the ventral visual pathway that is not in line with observed transitions in representational content along the anatomical posterior-to-anterior axis in human. Finally, they didn’t assess the time course with which model elements take part in object recognition. The proposed project aims at: examining the object space of Bao et al., 2020 in human, testing and finetuning their proposals based upon experiments that dissociate the relevant dimensions, and characterizing the temporal dynamics of the refined object representation model.

Date:1 Nov 2021 →  Today
Keywords:Object recognition, Deep neural networks, Neuroimaging
Disciplines:Cognitive neuroscience, Artificial intelligence, Knowledge representation and machine learning, Neurocognitive patterns and neural networks