6/By fusing EEG (temporal precision) with fMRI (spatial precision), this study offers a spatiotemporally resolved map of how audiovisual stimuli unfold across the brain.

5/ A two-branch deep neural network (DNN) model trained on audiovisual data couldn't replicate early cross-modal integration seen in the brain, emphasizing the need for early fusion in models

4/ High-level semantic & categorical info only appears later in high level auditory, visual, and multisensory areas.

3/ visual & auditory features were processed with similar onsets, but different temporal dynamics. Auditory info takes longer to peak than visual, likely due to the temporal nature of sounds.

2/ Using EEG and fMRI, we discovered early asymmetrical cross-modal interactions, with acoustic information represented in both early visual and auditory regions, while visual information only identified in visual cortices.

1/ The human brain seamlessly integrates distinct sensory information to create a coherent percept. This paper dives deep into where and when different types of audiovisual information are processed in the brain using naturalistic stimuli.