Detecting Road Surface Wetness from Audio: A Deep Learning Approach

TL;DR

This paper presents a deep learning model that detects road surface wetness from tire-surface interaction sounds, demonstrating high accuracy across diverse conditions and vehicle speeds.

Contribution

A novel recurrent neural network architecture for wetness detection from audio, validated on extensive real-world data with robust performance.

Findings

Achieved 93.2% unweighted average recall across all vehicle speeds.

Model maintains accuracy even at 0 mph due to environmental sound cues.

Robust performance across diverse road and environmental conditions.

Abstract

We introduce a recurrent neural network architecture for automated road surface wetness detection from audio of tire-surface interaction. The robustness of our approach is evaluated on 785,826 bins of audio that span an extensive range of vehicle speeds, noises from the environment, road surface types, and pavement conditions including international roughness index (IRI) values from 25 in/mi to 1400 in/mi. The training and evaluation of the model are performed on different roads to minimize the impact of environmental and other external factors on the accuracy of the classification. We achieve an unweighted average recall (UAR) of 93.2% across all vehicle speeds including 0 mph. The classifier still works at 0 mph because the discriminating signal is present in the sound of other vehicles driving by.