John F. Magnotti

The paper demonstrates that a deep neural network model, AVHuBERT, can replicate the McGurk effect, showing similar perceptual responses to humans when exposed to audiovisual speech stimuli.

This paper develops a machine learning model to identify seizure onset zones using iEEG data and predicts post-operative seizure freedom, achieving high accuracy in both tasks.

The paper applies the noisy encoding of disparity model to predict McGurk effect perception in Japanese speakers, showing high variability and accurate predictions.

YAEL streamlines electrode localization in iEEG by integrating all steps into a single GUI, enabling efficient and automated localization with high-quality visualization.

This paper introduces a method to compare lesion-behavior maps by predictive validity, showing it outperforms existing methods in detecting distinct lesion patterns for behaviors.

RAVE is an open-source software for analyzing and visualizing intracranial EEG data, enabling users to perform custom analyses and visualize results without coding.

This paper uses iEEG deconvolution to study neural responses to visual speech in human posterior superior temporal gyrus, finding subadditive multisensory processing and visual enhancement of auditory speech.

The paper uses the CIMS model to explain the relationship between the McGurk effect and everyday speech perception, showing stimulus-level correlations and observer-level discrepancies through causal inference and sensory encoding.

Dynamic electrical stimulation of visual cortex sites enables sighted and blind humans to perceive recognizable letter shapes without training.

The paper investigates tactile search behavior over the body using a tactile analog of the visual search task, finding serial search processes and frequency-dependent performance.

The paper shows that group differences in the McGurk effect are inflated due to high within-group variability and small sample sizes, leading to unreliable conclusions.