Please note, as of January 1st 2022, Person Detection will exclusively be available on Premium Service, click here to find out more.
Person Detection uses computer vision to detect people within your video. Through this feature, you can customize your video and alert preferences to decide whether you want to see video and alerts for all motion, people only, or have them suppressed.
Navigate to Modes and Motion Alerts within the app to customize your Canary experience.
For example, you can set your device to record only people, and send alerts of only people too. With this configuration setting, your timeline will be less cluttered, and you will receive fewer alerts as notifications won’t be sent for videos in which people are not detected.
Person Detection is only available on Canary devices that are running firmware version 3.0.0 or above.
Why are my pets sometimes detected as people?
Person detection uses advanced computer vision algorithms to identify people in your video and send you alerts. Sometimes, pets are incorrectly detected as people (vice versa). We’re continually tuning our algorithms to improve Person Detection accuracy, reducing false alerts, while providing you with security you can trust.
What happened to the motion notifications slider?
Since we introduced Person Detection in February 2018, users were able to control their notification experience by adjusting a slider with values from one to ten. For example, setting the slider to three or less would send alerts for people only, while anything above three would give you both person and general motion alerts.
In April 2018, we introduced mobile app version 2.15.0 making it easier to understand your settings and giving you more visibility into which notifications you’re receiving from Canary and why. Instead of a slider, there is now a collection of intuitive configuration options you can set to meet your specific needs.
What is the yellow rectangle on the timeline?
The yellow rectangle on the thumbnail of your timeline video represents where Canary's algorithms identified some type of movement triggered by a person, pet, or some other activity.