Publication detail

Acoustic insights: advancing object classification in urban landscapes using distributed acoustic sensing and convolutional neural networks

TOMAŠOV, A. BUKOVSKÝ, J. ZÁVIŠKA, P. HORVÁTH, T. LÁTAL, M. MÜNSTER, P.

Original Title

Acoustic insights: advancing object classification in urban landscapes using distributed acoustic sensing and convolutional neural networks

Type

conference paper

Language

English

Original Abstract

The paper introduces an innovative object classification method for urban environments, employing distributed acoustic sensing (DAS) to address the complexities of urban landscapes. Utilizing omnipresent optical telecommunication cables, our approach involves a modified convolutional neural network (CNN) with transfer learning, achieving up to 85% accuracy. This method reuses most of the original network for feature extraction, with a final layer customized for new urban datasets – initially trained at the Brno University of Technology and then adapted to city center data. The model effectively identifies urban elements like vehicles and pedestrians, showcasing the potential of DAS for real-time classification in urban management and planning.

Keywords

Convolutional neural networks;Machine learning;Cross validation;Fiber optics;Telecommunications;Fiber optics sensors

Authors

TOMAŠOV, A.; BUKOVSKÝ, J.; ZÁVIŠKA, P.; HORVÁTH, T.; LÁTAL, M.; MÜNSTER, P.

Released

18. 6. 2024

ISBN

0277-786X

Periodical

Proceedings of SPIE

Year of study

13017

Number

1301715

State

United States of America

Pages count

5

BibTex

@inproceedings{BUT188412,
  author="Adrián {Tomašov} and Jan {Bukovský} and Pavel {Záviška} and Tomáš {Horváth} and Michal {Látal} and Petr {Münster}",
  title="Acoustic insights: advancing object classification in urban landscapes using distributed acoustic sensing and convolutional neural networks",
  booktitle="Machine Learning in Photonics",
  year="2024",
  journal="Proceedings of SPIE",
  volume="13017",
  number="1301715",
  pages="5",
  doi="10.1117/12.3021990",
  issn="0277-786X"
}