TY - GEN
T1 - Participatory sound meter calibration system for mobile devices
T2 - 18th ACM Conference on Embedded Networked Sensor Systems, SenSys 2020
AU - Wu, Sheng Chun
AU - Wu, Dong Yi
AU - Ching, Fu Hsiang
AU - Chen, Ling Jyh
N1 - Funding Information:
This work was supported in part by the Ministry of Science and Technology of Taiwan, and in part by the Academia Sinica under Grant MOST 109-2221-E-001-005-MY3 and Grant AS-SS-109-02.
Publisher Copyright:
© 2020 ACM.
PY - 2020/11/16
Y1 - 2020/11/16
N2 - Noise exposure has been the emerging environmental factor for human health. Yet an accurate and large-scale sound monitoring network is not available due to the expense of high-quality professional sound level meters and poorly-calibrated low-cost noise sensors. In this work, we propose a participatory sound meter calibration using smartphones. The system employs a low-cost and open-sourced calibration station to conduct side-by-side sound measurements, and all the measurement data are uploaded to the open data portal to build calibration models for different phone brands and models. We show that, using our calibration models, the MAE of calibration performance can be reduced significantly from 12.4 dbA to 2.8 dbA for the same device and 3.3 dbA for the other device of the same phone model. The results of this study can benefit crowdsourcing-based large-scale sound measurements and facilitate noise exposure, public health, and smart city researches in the future.
AB - Noise exposure has been the emerging environmental factor for human health. Yet an accurate and large-scale sound monitoring network is not available due to the expense of high-quality professional sound level meters and poorly-calibrated low-cost noise sensors. In this work, we propose a participatory sound meter calibration using smartphones. The system employs a low-cost and open-sourced calibration station to conduct side-by-side sound measurements, and all the measurement data are uploaded to the open data portal to build calibration models for different phone brands and models. We show that, using our calibration models, the MAE of calibration performance can be reduced significantly from 12.4 dbA to 2.8 dbA for the same device and 3.3 dbA for the other device of the same phone model. The results of this study can benefit crowdsourcing-based large-scale sound measurements and facilitate noise exposure, public health, and smart city researches in the future.
KW - crowdsourcing
KW - smartphone
KW - sound meter calibration
UR - http://www.scopus.com/inward/record.url?scp=85097549231&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85097549231&partnerID=8YFLogxK
U2 - 10.1145/3384419.3430449
DO - 10.1145/3384419.3430449
M3 - Conference contribution
AN - SCOPUS:85097549231
T3 - SenSys 2020 - Proceedings of the 2020 18th ACM Conference on Embedded Networked Sensor Systems
SP - 709
EP - 710
BT - SenSys 2020 - Proceedings of the 2020 18th ACM Conference on Embedded Networked Sensor Systems
PB - Association for Computing Machinery, Inc
Y2 - 16 November 2020 through 19 November 2020
ER -
