Unsupervised vehicle recognition using incremental reseeding of acoustic signatures

TL;DR

This paper presents an unsupervised method for vehicle recognition from roadside audio using spectral embedding and incremental reseeding, enabling accurate identification of vehicles based on their acoustic signatures without labeled data.

Contribution

It introduces an unsupervised approach combining spectral embedding and incremental reseeding for vehicle recognition from audio signals, advancing beyond traditional supervised methods.

Findings

Incremental reseeding accurately identifies vehicles from acoustic data.

Spectral embedding effectively reduces dimensionality of audio signatures.

Unsupervised method performs well without labeled training data.

Abstract

Vehicle recognition and classification have broad applications, ranging from traffic flow management to military target identification. We demonstrate an unsupervised method for automated identification of moving vehicles from roadside audio sensors. Using a short-time Fourier transform to decompose audio signals, we treat the frequency signature in each time window as an individual data point. We then use a spectral embedding for dimensionality reduction. Based on the leading eigenvectors, we relate the performance of an incremental reseeding algorithm to that of spectral clustering. We find that incremental reseeding accurately identifies individual vehicles using their acoustic signatures.