Object representations in the human brain reflect the co-occurrence statistics of vision and language

This study investigates whether object co-occurrence statistics in visual environments and language are encoded in the human visual system. Using machine learning and fMRI, the researchers found that cortical responses to single objects are predicted by the statistical ensembles in which they typically appear. This link between objects and their visual contexts is strongest in the parahippocampal cortex, overlapping with the scene-selective parahippocampal place area. A language-based statistical model predicted responses in neighboring object-selective visual cortex regions. The findings suggest that object sensory coding reflects the latent statistics of object context in visual and linguistic experience.

Michael F. Bonner, Russell A. Epstein