Automatic Conflict Detection in Police Body-Worn Audio

TL;DR

This paper presents a new audio processing pipeline that improves conflict detection in police body-worn audio by using adaptive noise removal, non-speech filtering, and novel conflict measures based on speech repetition and intensity.

Contribution

The study introduces a tailored pipeline with innovative conflict metrics that outperform traditional methods in noisy, real-world police audio scenarios.

Findings

Effective conflict detection on police body-worn audio

Improved metrics based on phrase repetition and intensity

Demonstrated on real LAPD data

Abstract

Automatic conflict detection has grown in relevance with the advent of body-worn technology, but existing metrics such as turn-taking and overlap are poor indicators of conflict in police-public interactions. Moreover, standard techniques to compute them fall short when applied to such diversified and noisy contexts. We develop a pipeline catered to this task combining adaptive noise removal, non-speech filtering and new measures of conflict based on the repetition and intensity of phrases in speech. We demonstrate the effectiveness of our approach on body-worn audio data collected by the Los Angeles Police Department.