Road Induced Noise Prediction using Deep Learning from Road Surface Images

Abstract

The purpose of this research is to predict both the frequency and peak level of road induced noise from the signals of monocular camera adopted for automatic driving without newly adding sensors for utilizing the prediction result as a reference signal to reduce road induced noise by ANC. Using the monocular camera, both the frequency and peak level of road induced noise is predicted by machine learning from the road surface features which are extracted from road surface images by three image recognition techniques (HOG, CNN, autoencoder).