System and Method for Identifying a Vehicle via Audio Signature Matching

A computer-implemented method includes providing a first application to a mobile computing device that obtains a set of audio signals, transforms the set of audio signals into a set of signal features, wherein no audio signal in the set of audio signals can be derived from the set of signal features, and transmits the set of signal features to a remote computing system, without transmitting any portion of the set of audio signals. The method also includes receiving, at the remote computing system, the set of signal features, identifying a type of the vehicle at least in part by applying the set of signal features to a trained neural network, and one or both of (i) causing the identified type of the vehicle to be displayed, and (ii) causing the identified type of the vehicle to be provided to a second application.