Nearly 50% of police officers wear body cameras, but while hundreds of hours of footage is recorded per day at a department, only a fraction is ever analyzed and reviewed. As many departments look for better oversight and training of their police force, few are able to leverage body camera data as a source of insight into their interactions with the community.
Research shows that language used in police interactions as measured by humans reviewing body worn camera video shows disparities in officer behavior based on the use of respectful or disrespectful language.
Photo taken from PNAS, Language from police body camera footage shows racial disparities in officer respect.
Simply put: using more respectful language leads to fewer escalated scenarios. However, the vast amount of footage to be reviewed to make determinations of the use of respectful language across a department is nearly impossible to process with solely human review.
Truleo transcribes BWC audio and separates the audio into individualized, anonymous speakers. The speaker wearing the camera is tagged anonymously as the officer. Truleo’s NLP models then run on the speaker-separated transcript, identifying key phrases associated with risky or respectful interactions. Features are weighted based on a department’s preference for detection (e.g. directed profanity is worse than informality). In addition, Truleo’s models tag events, like arrests and use of force, as a further dimension to analyze risk.
Truleo’s Natural Language Processing models use a modern architecture called a “transfomer”. The key advance is that these models learn based on context, not keywords. Thus, seeing a word used in context, the model can automatically extrapolate synonyms or other potential variations of the word. In this way, Truleo’s models are able to capture the key phrases associated with risk and respect with only a handful of examples.
Examples are shown below of the types of phrases Truleo’s models use to determine respect or risk: