About the Project
How do we understand what we see? Visual object recognition is known to rely on the ventral visual pathway through the occipital and temporal lobe in the brain. Successful recognition requires the transformation of neural activity to access the meaning of the visual image. This happens with incredible speed, where semantic information can be read-out from the brain within half a second of seeing an image. The aim of this project is to better understand the brain mechanisms that rapidly transform…
visual inputs into meaningful representations.
We know that the brain processes lots of different forms of information about objects, and these change over time. One approach to understand brain representations is to map objects onto different positions within a multidimensional space. Two objects with similar brain activity will be in a similar position, while objects with different activity will be in different positions. For example, visual regions will represent a red apple and cricket ball in a similar way, but a banana and apple very differently, while semantic regions will group the banana and apple, and not the apple and cricket ball. Over time, the position of objects move in this space, from a visual to semantic representation, creating a trajectory through a phase space. This project aims to decompose these phase spaces to understand the reasons objects cluster in different ways over time. Doing so, will provide fundamental mechanistic insights into how the brain supports cognition.
The project will use time-sensitive methods such as MEG and EEG which measure neural activity in the human brain with high temporal resolution. Data analyses will use various methods such as multivariate pattern analysis, neural decoding and phase-space analysis. In addition to lab-based brain imaging studies, the project has scope to use mobile EEG recordings and mixed reality technology to investigate object processing beyond the lab, in real-world settings.
