Facial Recognition Technology

Why did York Regional Police partner with Peel Regional Police to acquire facial recognition technology?

The partnership between York and Peel Regional Police will increase investigative capacity by increasing the size of the database of images. The partnership also decreases the cost to purchase, develop, maintain and operate the facial recognition system.

Police know that criminals don't adhere to jurisdictional borders—this shared system encourages collaboration and inter-agency information sharing to make our communities safer.

What is the facial recognition system? How does it work?

When police collect images of suspects or persons of interest during the course of a criminal investigation, they can use the facial recognition system to compare faces in those images with faces in the criminal booking image (mugshot) database. This database comprises images collected lawfully under the Identification of Criminals Act.

Facial recognition technology is only one tool used to generate investigative leads, which will need to be further corroborated with other independent evidence and investigative techniques.

Comparisons made using the system will not stand alone in the judicial process and cannot be used alone to identify, charge or convict a suspect.

Facial recognition technology protects the integrity of the investigative process as it is not susceptible to the biases that exist in eyewitness accounts.

The facial recognition system will assist officers to identify suspects more quickly and more accurately, freeing up invaluable investigative resources. It will drastically reduce time spent manually comparing images to mugshots.

Facial recognition technology assists with identifying potential suspects or persons of interest while eliminating others from an investigation.

This technology represents a natural step forward for police investigations and will serve as another tool in our efforts to keep all people in our regions safe.

Yes. Law enforcement agencies across Canada are already using facial recognition technology to help solve crimes, including homicides, sexual assaults, robberies and others, in the communities they serve.

Facial recognition technology is already deployed at airports and other border checkpoints across the country to keep Canadians safe.

The system does not violate citizens’ privacy.

Police have a legislative duty to investigate criminal activities while safeguarding the fundamental rights guaranteed by the Canadian Charter of Rights and Freedoms and the Ontario Human Rights Code. Police in York and Peel Regions recognize their responsibility to the communities they serve and will ensure they provide the best investigative capabilities possible.  

Police will not be collecting images from any live streaming footage or randomly searching images. 

Our members will receive training and be subject to stringent operating procedures which outline their roles and responsibilities surrounding the ethical use of facial recognition.

Data collected is well protected against loss, hacking, modification and unauthorized access.

York Regional Police will continue to ensure the highest standards in the collection of data and best practices and policies around use of the technology are adhered to.

Why now?

In the past five years, facial recognition technology has made tremendous strides in accuracy and demographic differences according to the National Institute of Standards and Technology (NIST). York Regional Police and Peel Regional Police waited until the technology improved to pursue its implementation.

It is important that police remain vigilant and keep current with the latest technologies and innovations in order to prevent criminal activities and keep our communities safe. Our licensing agreement with IDEMIA ensures police have access to the most up-to-date, highest-performing technology available.

All investigative leads driven by facial recognition technology are independently verified and evaluated by forensic facial recognition analysts.

Is facial recognition technology biased, or is it accurate?

Research has shown that early facial recognition algorithms were prone to varying demographic differences. However, the exact differences depended on the developer of the algorithm and type of facial recognition application.

As of 2022, IDEMIA’s facial recognition technology ranked most accurate on the false match rate (FMR) fairness test among the 100 most accurate algorithms in the latest Face Recognition Vendor Test (FRVT) from the National Institute of Standards and Technology (NIST).

IDEMIA’s false match rate between different demographic groups hardly differs. The algorithm identifies all subjects equally well, regardless of demographic.

Research now indicates that facial recognition technology achieves near perfect accuracy when combined with human adjudication in a law enforcement application.

For more information visit the IDEMIA website.

Individuals whose criminal booking image (mugshot) and fingerprints have been gathered by York Regional Police pursuant to the Identification of Criminals Act, may be eligible to have their image and fingerprints destroyed, provided certain criteria are met. Any fees associated with requests for image and fingerprint destruction are being waived by York Regional Police.