Dataset - Adversarial AutoEncoder and Multi-Armed Bandit for Dynamic Difficulty Adjustment in Immersive Virtual Reality for Rehabilitation: Application to Hand Movement

This dataset consists of movements drawn with a wireless remote controller in an immersive VR environment for 6 different movements called "Cube," "Cylinder", "Heart", "Infinity", "Sphere" and "Triangle".
Data were collected from 10 participants. Each movement was collected 3 times for each participant for each session and 3 sessions were performed thus providing a total of 9 repetitions of each movement per participant.
The data were projected onto the frontal plane facing the head mounted display. The total number of movements collected in the database is 540.
Data were collected using Unity 2020.3.26f1 with the Oculus Rift S and resampled to 32 points.

A Python script is also included with an example of how to load the data. The Python script was tested with:
# Python - 3.8.8
# Panda - 1.3.1
# Numpy - 1.21.1
# Matplotlib - 3.4.2