The generative capabilities of the human visual system

Much of life is spent imagining or dreaming of internal images that one has never actually observed. Why is the human visual system so good at generating images, and how does this remarkable ability help us to see? We are addressing this question by monitoring the human brain as it engages complex, real-world scenery and as it calls upon memory to generate mental images. Using a combination of functional magnetic resonance imaging (fMRI), human intracranial EEG (iEEG), and mathematical modeling we aim to reveal the functional role of generative vision.

Neural decoding

What is the most efficient algorithm for converting human brain activity into a picture one’s visual experience? We are developing methods for decoding information about subjective visual experience from multiple stages of the visual system. For technical details see the many decoding methods presented in our recent papers (Kay et al., 2008; Naselaris et al., 2009; Nishimoto et al., 2011; Naselaris et al., 2012; Stansbury et al., 2013).

Modeling the brain’s response to the natural environment

Natural visual environments are a cluttered mess of spatially overlapping objects. How does the visual system identify, categorize, and localize the multitude of objects that crowd it’s visual field? We are developing predictive models that characterize visual responses to the many distinct kinds of information present in natural visual environments. For a review of these modeling efforts see Naselaris et al., 2011.