Abstract

This dataset contains inertial measurements data (accelerometer, gyroscope and magnetometer), recorded as a part of DoorINet research. Our dataset was recorded using two types of IMUs: Memsense MS-IMU3025 [32] and Movella Xsens DOT [33]. The Memsense MS-IMU3025 was used to generate the ground-truth (GT) readings. This IMU has a gyroscope bias instability of 0.8°/h over the axis of interest Z and was recorded at 250Hz. The Movella Xsens DOT IMUs were used as units under test. It has a gyroscope bias instability of 10°/h over the axis of interest Z and was recorded at 120Hz.

Inertial sensors were placed on a door surface, and the door was opened and closed, inertial sensors registered door movements.

The table structure is as follows:

* _sampletimefine_ - internal time from the IMU, in microseconds;

* _acc_x_ - accelerometer measurements along the X axis, in g;

* _acc_y_ - accelerometer measurements along the Y axis, in g;

* _acc_z_ - accelerometer measurements along the Z axis, in g;

* _gyr_x_ - gyroscope measurements along the X axis, in degrees per second;

* _gyr_y_ - gyroscope measurements along the Y axis, in degrees per second;

* _gyr_z_ - gyroscope measurements along the Z axis, in degrees per second;

* _mag_x_ - magnetometer measurements along the X axis, in nano Tesla;

* _mag_y_ - magnetometer measurements along the Y axis, in nano Tesla;

* _mag_z_ - magnetometer measurements along the Z axis, in nano Tesla;

* _heading_angle_ - resulting heading angle of the door opening (ground truth), in degrees;

* _session_ - recording session number, either 1, 2 or 3;

* _xdot_label_ - label of the XDOT IMU, we had XDOT IMUs labelled 5, 6, 7, 8, 9, 10, 11, 12, 14 and 15 (see photos above);

* _xdot_position_ - position of the XDOT IMU on the door surface, can be either "handle" (close to the door handle), "median" (in the middle of the door) and "hinge" (close to the door hinge);

* _description_ - some textual description of the current experiment

Instructions:

The datasets are written to CSV files and are organized as follows:

* **train_dataset.csv** contains all the training IMU data used in the paper;

* **unused_data.csv** contains IMU data that was recorded but was not used in training for various reasons;

* **test_12.csv** is a test dataset recorded by XDOT IMU #12;

* **test_14.csv** is a test dataset recorded by XDOT IMU #14;

* **test_5.csv** is a test dataset recorded by XDOT IMU #5;