Why we need more AI in public safety, not less
By John Newsom
While public safety initiatives have made great strides in protecting civilians, innovative artificial intelligence technology is the next step that cities should take to protect their civilians.
For example, while critics of police surveillance often say it’s a breach of privacy, there are plenty of crimes that have been solved in large part by review of CCTV footage, such as the Boston Marathon bombing. The suspects were not only found within 24 hours, but CCTV footage enabled immediate communication with the public about who and where to be on the lookout in case of suspicious activity.
Without valuable AI insights, public safety and police officials are often unable to sort and react to critical and valuable audio/video information regarding a case or trial.
Misunderstanding creates fear
Some hesitancy surrounding artificial intelligence comes from the fact that it’s new and not generally understood. In order to better protect cities, an understanding of what is possible (and what is not) should be shared. When discussing AI, “engines” are the types of systems used to accomplish a task. For example:
Transcription engines utilize natural language processing (NLP) to analyze audio streams from digital audio recorders, wiretaps, body and dash cameras, security cameras, and interview room recordings to find any word spoken and to produce transcriptions in all major languages.
Face detection engines identify the presence of an anonymous face in video, and correlate that face across multiple video sources. Face detection can be augmented for facial identification by utilizing a reference library of known people, and can also be extended to the redaction of anonymous faces other than a key subject. Some face engines detect and describe emotions, such as anger or fear.
Object identification engines can be used to identify the make and model of vehicles, and to identify weapons, unattended luggage, or other specialized tasks.
Audio/video fingerprinting engines identify segments of media that are a match to certain sound profiles. These engines use specialized libraries of reference audio clips, such as gunshots, sirens, or specific phrases used in law enforcement.
Motion detection engines can identify specific actions such as perpetrators fighting, stealing, or running suddenly.
Geolocation engines compile and track coordinates and time synchronization across multiple video sources such as surveillance cameras. Similarly, Landmark identification engines can be trained to recognize certain landmarks within a municipality.
Data will continue to grow
These AI engines can be used to tackle one of the biggest problems involving technology and law enforcement: In times of investigation or crisis, the amount of data surrounding a specific case may feel insurmountable. Mary Fan, a law professor at the University of Washington, claims that law enforcement agencies in large cities are creating 10,000 hours of video a week , an impossible amount of information for a human to work through.
Artificial intelligence can rapidly ingest, analyze and organize all of this data, which will save thousands of hours of manual searching, giving agencies the opportunity to focus time and resources on more mission critical tasks. While the push to create unique and modern solutions to tackle public safety issues can be exciting, a method must exist to amalgamate the information collected from these disparate sources into a single IT platform. Only then can a technology platform deliver a comprehensive picture of urban public safety and facilitate communication and collaboration among all concerned parties.
John Newsom is the executive vice president and GM for Veritone Government, an artificial intelligence company.