CNN architectures for road surface wetness classification from acoustic signals

The classification of road surface wetness is important for both the development of future driverless vehicles and the development of existing vehicle active safety systems. Wetness on the road surface has an impact on road safety and is one of the leading causes of weather-related accidents. Alth...

Full description

Bibliographic Details
Main Authors: Bahrami, Siavash, Doraisamy, Shyamala, Azman, Azreen, Nasharuddin, Nurul Amelina, Yue, Shigang
Format: Article
Language:English
Published: Springer
Online Access:http://psasir.upm.edu.my/id/eprint/104489/
http://psasir.upm.edu.my/id/eprint/104489/1/CNN%20Architectures%20for%20Road%20Surface%20Wetness%20Classification%20from%20Acoustic%20Signals.pdf
_version_ 1848864304933109760
author Bahrami, Siavash
Doraisamy, Shyamala
Azman, Azreen
Nasharuddin, Nurul Amelina
Yue, Shigang
author_facet Bahrami, Siavash
Doraisamy, Shyamala
Azman, Azreen
Nasharuddin, Nurul Amelina
Yue, Shigang
author_sort Bahrami, Siavash
building UPM Institutional Repository
collection Online Access
description The classification of road surface wetness is important for both the development of future driverless vehicles and the development of existing vehicle active safety systems. Wetness on the road surface has an impact on road safety and is one of the leading causes of weather-related accidents. Although machine learning algorithms such as recurrent neural networks (RNN), support vector machines (SVM), artificial neural networks (ANN) and convolutional neural networks (CNN) have been studied for road surface wetness classification, the improvement of classification performances are still widely being investigated whilst keeping network and computational complexity low. In this paper, we propose new CNN architectures towards further improving classification results of road surface wetness detection from acoustic signals. Two CNN architectures with differing layouts for its dropout layers and max-pooling layers have been investigated. The positions and the number of the max-pooling layers were varied. To avoid overfitting, we used a 50% dropout layers before the final dense layers with both architectures. The acoustic signals of tyre to road interaction were recorded via mounted microphones on two distinct cars in an urban environment. Mel-frequency cepstral coefficients (MFCCs) features were extracted from the recordings as inputs to the models. Experimentation and comparative performance evaluations against several neural networks architectures were performed. Recorded acoustic signals were segmented into equal frames and thirteen MFCCs were extracted for each frame to train the CNNs. Results show that the proposed CMCMDD1 architecture achieved the highest accuracy of 96.36% with the shortest prediction time.
first_indexed 2025-11-15T13:46:42Z
format Article
id upm-104489
institution Universiti Putra Malaysia
institution_category Local University
language English
last_indexed 2025-11-15T13:46:42Z
publisher Springer
recordtype eprints
repository_type Digital Repository
spelling upm-1044892023-08-18T03:05:42Z http://psasir.upm.edu.my/id/eprint/104489/ CNN architectures for road surface wetness classification from acoustic signals Bahrami, Siavash Doraisamy, Shyamala Azman, Azreen Nasharuddin, Nurul Amelina Yue, Shigang The classification of road surface wetness is important for both the development of future driverless vehicles and the development of existing vehicle active safety systems. Wetness on the road surface has an impact on road safety and is one of the leading causes of weather-related accidents. Although machine learning algorithms such as recurrent neural networks (RNN), support vector machines (SVM), artificial neural networks (ANN) and convolutional neural networks (CNN) have been studied for road surface wetness classification, the improvement of classification performances are still widely being investigated whilst keeping network and computational complexity low. In this paper, we propose new CNN architectures towards further improving classification results of road surface wetness detection from acoustic signals. Two CNN architectures with differing layouts for its dropout layers and max-pooling layers have been investigated. The positions and the number of the max-pooling layers were varied. To avoid overfitting, we used a 50% dropout layers before the final dense layers with both architectures. The acoustic signals of tyre to road interaction were recorded via mounted microphones on two distinct cars in an urban environment. Mel-frequency cepstral coefficients (MFCCs) features were extracted from the recordings as inputs to the models. Experimentation and comparative performance evaluations against several neural networks architectures were performed. Recorded acoustic signals were segmented into equal frames and thirteen MFCCs were extracted for each frame to train the CNNs. Results show that the proposed CMCMDD1 architecture achieved the highest accuracy of 96.36% with the shortest prediction time. Springer Article PeerReviewed text en http://psasir.upm.edu.my/id/eprint/104489/1/CNN%20Architectures%20for%20Road%20Surface%20Wetness%20Classification%20from%20Acoustic%20Signals.pdf Bahrami, Siavash and Doraisamy, Shyamala and Azman, Azreen and Nasharuddin, Nurul Amelina and Yue, Shigang CNN architectures for road surface wetness classification from acoustic signals. Lecture Notes in Electrical Engineering, 835. pp. 777-788. ISSN 1876-1100; ESSN: 1876-1119 https://link.springer.com/chapter/10.1007/978-981-16-8515-6_59 10.1007/978-981-16-8515-6_59
spellingShingle Bahrami, Siavash
Doraisamy, Shyamala
Azman, Azreen
Nasharuddin, Nurul Amelina
Yue, Shigang
CNN architectures for road surface wetness classification from acoustic signals
title CNN architectures for road surface wetness classification from acoustic signals
title_full CNN architectures for road surface wetness classification from acoustic signals
title_fullStr CNN architectures for road surface wetness classification from acoustic signals
title_full_unstemmed CNN architectures for road surface wetness classification from acoustic signals
title_short CNN architectures for road surface wetness classification from acoustic signals
title_sort cnn architectures for road surface wetness classification from acoustic signals
url http://psasir.upm.edu.my/id/eprint/104489/
http://psasir.upm.edu.my/id/eprint/104489/
http://psasir.upm.edu.my/id/eprint/104489/
http://psasir.upm.edu.my/id/eprint/104489/1/CNN%20Architectures%20for%20Road%20Surface%20Wetness%20Classification%20from%20Acoustic%20Signals.pdf