Enabling reproducible research in sensor-based transportation mode recognition with the Sussex-Huawei dataset
- Submitting institution
-
University of Sussex
- Unit of assessment
- 12 - Engineering
- Output identifier
- 335131_81170
- Type
- D - Journal article
- DOI
-
10.1109/ACCESS.2019.2890793
- Title of journal
- IEEE Access
- Article number
- -
- First page
- 10870
- Volume
- 7
- Issue
- -
- ISSN
- 2169-3536
- Open access status
- Compliant
- Month of publication
- January
- Year of publication
- 2019
- URL
-
https://doi.org/10.1109/ACCESS.2019.2890793
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- No
- Number of additional authors
-
5
- Research group(s)
-
-
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This paper is seminal in that it sets out how the research community ought to benchmark locomotion/transportation mode recognition systems from mobile devices using machine learning, to enable reproducible research in the wearable, mobile and ubiquitous computing communities. It does so by setting out twelve reference recognition scenarios, defining standardized combinations of sensors, defining precise evaluation criteria including train/test sets, and presenting benchmark results on the world’s largest dataset of mobile sensor data collected from mobile users, which totals 2812 hours of labelled data and 17562 km of traveled distance, and extremely precise annotations of locomotion and transportation modes.
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -