Abstract

Noise control on construction sites is essential to manage project risks including safety accidents, worker's health, and civil complaints. This paper presents two advanced Audio Spectrogram Transformer-based models to identify noise source types and estimate their respective noise levels on complex construction environments. The classification model first differentiated the noise sources from mixed noise. Receiving the classification results, the regression model then estimated their noise levels. Nine categories of on-site noise (i.e., six construction and three external noises) were collected, a total of 234.4-min audio files, and mixed randomly with different ratios for the validation purpose. As a result, the models successfully classified the noise sources with F1-score of 0.901 and estimated the respective noise levels with an average absolute error of 0.573 dB. This study is expected to clarify the contributors of the measured noise levels, enabling strategic noise management and sophisticated feedback from on-site noise data.

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.