Datasets:

Tachintech
/

Tactile_Gesture_Language

This dataset contains data of contact gestures. The goal of this dataset is to enable stable recognition of contact gestures on a flat surface. It was collected using a 32×32 sensor array on a plane, covering ten gesture types: stroking, tapping, pressing, scratching, pinching, poking, grabbing, squeezing, kneading, and beatting. The dataset includes data from 10 participants, with each participant contributing a different total amount of data.

Uses

An example data loading program is provided as load_sample.py, offering a straightforward way to load the data in NumPy format.

Dataset Structure

The dataset comprises data collected from ten participants. For each participant, all data are stored in a dedicated folder, and inside that folder, the files are organized by gesture class (one file per gesture).

Data Collection and Processing

The collected data were post-processed with clutter removal and manually checked for quality.