OxWearables / capture24

Capture24: Human activity recognition with activity trackers
Other
20 stars 6 forks source link

Activity recognition with the Capture-24 dataset

Check out tutorial.py or tutorial.ipynb

The Capture-24 dataset can be downloaded here

To run the examples, you will need numpy, pandas, sklearn, imblearn and tqdm. Most of these come with anaconda.

References

Dataset description and benchmark paper:

@misc{chan2024capture24,
      title={CAPTURE-24: A large dataset of wrist-worn activity tracker data collected in the wild for human activity recognition}, 
      author={Shing Chan and Hang Yuan and Catherine Tong and Aidan Acquah and Abram Schonfeldt and Jonathan Gershuny and Aiden Doherty},
      year={2024},
      eprint={2402.19229},
      archivePrefix={arXiv},
      primaryClass={cs.HC}
}

Papers featuring the Capture-24 dataset:

Walmsley R, Chan S, Smith-Byrne K, Ramakrishnan R, Smith-Byrne K, Woodward M, Rahimi K, Dwyer T, Bennett D, Doherty A (2021) Reallocating time from device-measured sleep, sedentary behaviour or light physical activity to moderate-to-vigorous physical activity is associated with lower cardiovascular disease risk. British Journal of Sports Medicine doi: 10.1136/bjsports-2021-104050

Gershuny J, Harms T, Doherty A, Thomas E, Milton K, Kelly P, Foster C (2020) Testing self-report time-use diaries against objective instruments in real time. Sociological Methodology doi: 10.1177/0081175019884591

Doherty A, Smith-Bryne K, Ferreira T, Holmes MV, Holmes C, Pulit SL, Lindgren CM (2018) GWAS identifies 14 loci for objectively-measured physical activity and sleep duration with causal roles in cardiometabolic disease. Nature Communications. 9(1):5257

Willetts M, Hollowell S, Aslett L, Holmes C, Doherty A. (2018) Statistical machine learning of sleep and physical activity phenotypes from sensor data in 96,220 UK Biobank participants. Scientific Reports. 8(1):7961

License

See license before using these materials.