- What is it?
HOOD - Human Odometry Outdoor Data Set is a public collection of labelled accelerometer and gyroscope data recordings to be used for the design and validation of Human Odometry (HO) systems. HOOD is specifically designed to address the following questions:
- which is the optimal sensor placement for a HO system?
- what is the performance of human odometry algorithms for uncommon motions such as crawling and slithering?
- what is the performance of human odometry algorithms for uncommon grounds such as river bed, woods and snow?
- is it possible to design a single, robust and inexpensive HO system that can reliably reconstruct the followed path for different motion types and path configurations, based on accelerometer and gyroscope information exclusively?
HOOD is composed of 168 recordings of accelerometer and gyroscope data collected from one volunteer and referring to the combinations of 6 motion types (slow walking, normal walking, running, slow crawling, fast crawling, slithering), 6 outdoor environments (grass field, uphill road, staircase, river bed, woods, snow), 4 sensor placements (foot, waist, wrist, chest) and 2 different path configurations (straight line, zig-zag).
An article describing the dataset is currently under submission at IROS 2015.
- Version
Version: 2.0
Released on: 17/03/2015
17/03/2015: version 2.0 is published, with an extended dataset
25/11/2014: version 1.1 is published, with clearer data organization
28/09/2014: version 1.0 is published
- Documentation
Up-to-date documentation for this release is provided in the file MANUAL.pdf
- Installation & usage
This dataset does not require any installation.
The provided MATLAB script "viewRawData.m" allows for visualization of the recorded accelerometer data. A description and example usage of the script can be accessed within MATLAB environment with the command: help viewRawData
The provided MATLAB script has been developed and tested with MATLAB R2013b.
- Licensing
This dataset is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY, including the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. The authors allow the users of the "HOOD - Human Odometry Outdoor Data Set" to use and modify it for their own research. Any commercial application, redistribution, etc... has to be arranged between users and authors individually.
For further license information, please contact the authors.
- Authors contacts
If you want to be informed about dataset updates and new code releases, obtain further information about the provided dataset, or contribute to its development please write to:
-
Barbara Bruno, dept. DIBRIS, Università degli Studi di Genova (Italy) barbara.bruno@unige.it
-
Fulvio Mastrogiovanni, dept. DIBRIS, Università degli Studi di Genova (Italy) fulvio.mastrogiovanni@unige.it
-
Antonio Sgorbissa, dept. DIBRIS, Università degli Studi di Genova (Italy) antonio.sgorbissa@unige.it
Barbara Bruno, Fulvio Mastrogiovanni and Antonio Sgorbissa are with DIBRIS (Department of Computer Engineering, Bioengineering, Robotics and Systems Engineering) at the University of Genova, via Opera Pia 13, 16145, Genova, Italia