A dataset of throat-related events collected with a soft skin-attachable vibration sensor

This dataset is collected using a skin-attachable sensor placed on the laryngeal prominence. The sensor measures various throat-related events, including coughing, speaking, swallowing, and throat clearing. The dataset is divided into a training set and a test set, each containing different types and quantities of events.

Training set

The training set comprises the following throat-related events:

1. Coughing: 1,984 instances2. Speaking (English): 2,107 instances3. Swallowing: 2,028 instances4. Throat Clearing: 2,089 instances

Test set

The test set includes a diverse range of events to evaluate the model's performance:

1. Coughing: 150 instances2. Speaking (multiple languages): 333 instances (comprising English, French, German, Korean, and Spanish)3. Swallowing: 161 instances4. Throat Clearing: 148 instances

This dataset offers labeled throat-related events for training and testing machine learning models.