Chewing event detection dataset: 3D facial landmarks and sEMG-derived labels

This dataset contains processed frame-wise data associated with the study “A video-based measurement framework for chewing event detection using 3D facial landmark dynamics and sEMG-based reference annotation”.

The release includes 3D facial landmark sequences extracted from frontal videos using MediaPipe Face Mesh and synchronized binary chewing labels derived from bilateral anterior temporalis surface electromyography. Each CSV file corresponds to one acquisition/session and contains 468 landmarks with x, y, and z coordinates, plus a binary chewing_label column. Rows correspond to video frames in temporal order.

The dataset is intended to support reproducibility of the preprocessing, temporal-window generation, and chewing-event detection workflow described in the associated manuscript. Raw facial videos are not included for privacy reasons. Example Python code is provided to load the CSV files, reshape the landmark coordinates, extract the labels, and generate causal temporal windows.