HIAP - Human-Inspired Artificial Perception

The research activity aims to investigate the neural and computational principles that structure object representations in the human visual systems. Through an integrated cognitive-computational approach, the project will acquire functional neuroimaging data (fMRI) to map how different behavioural goals influence neural representations of objects and their spatial organization. These data will be analysed using advanced multivariate techniques, including Representational Similarity Analysis (RSA), to characterise the representational geometry underlying visual object processing. In addition, state-of-the-art AI models of vision will be used to model and quantify the structure of visual representations, enabling direct comparisons between neural, behavioural, and computational representational spaces.