DoorInet_dataset

Citation Author(s):
Aleksei
Zakharchenko
University of Haifa
Submitted by:
Aleksei Zakharchenko
Last updated:
Wed, 07/17/2024 - 00:19
DOI:
10.21227/jzf5-6e97
Data Format:
*.csv
License:
Creative Commons Attribution
55 Views
Categories:
Digital signal processing
Machine Learning
Keywords:
indoor positioning; pedestrian dead reckoning; inertial measurement unit; accelerometer; gyroscope;
0
0 ratings - Please login to submit your rating.
ACCESS DATASET CITE SHARE/EMBED

Abstract 

Inertial sensors are widely used in a variety of applications. A common task is orientation estimation. To tackle such a task, attitude and heading reference system algorithms are applied. Relying on the gyroscope readings, the accelerometer measurements are used to update the attitude angles, and magnetometer measurements are utilized to update the heading angle. In indoor environments, magnetometers suffer from interference that degrades their performance. It mainly impacts applications that estimate the heading angle of moving objects, such as walking pedestrians, closets, and refrigerators. To circumvent such situations, we propose DoorINet, an end-to-end deep-learning framework to calculate the heading angle from door-mounted, low-cost inertial sensors without using magnetometers. To evaluate our approach, we record a unique dataset containing 391 minutes of accelerometer and gyroscope measurements and corresponding ground-truth heading angle. We show that our proposed approach outperforms commonly used,model based approaches and data-driven methods. To enable reproducibility of our results and future research, both code and data are available at https://github.com/ansfl/DoorINet

Instructions: 

File structure is as follows:

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

* 12_test.xlsx is a test dataset recorded by XDOT IMU #12;

* 14_test.xlsx is a test dataset recorded by XDOT IMU #14;

* 5_test.xlsx is a test dataset recorded by XDOT IMU #5;

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

Dataset Files

LOGIN TO ACCESS DATASET FILES

 

QUESTIONS?