I run Frigate to record security cameras and detect people, cars, and animals when in view. My current Frigate server runs on a Raspberry Pi CM4 and a Coral TPU plugged in via USB.
Raspberry Pi offers multiple AI HAT+'s for the Raspberry Pi 5 with built-in Hailo-8 or Hailo-8L AI coprocessors, and they're useful for low-power inference (like for image object detection) on the Pi. Hailo coprocessors can be used with other SBCs and computers too, if you buy an M.2 version.

Jeff Geerling is a software developer, author, and open-source contributor known for his expertise in infrastructure automation and Drupal development. Readers can learn about DevOps practices, Ansible automation, and web development with Drupal. With blog posts, tutorials, and open-source projects, Jeff Geerling shares  insights and best practices for building and managing web infrastructure.

Jeff Geerling

A practical guide to setting up Frigate NVR with a Hailo AI coprocessor on a Raspberry Pi 5 for hardware-accelerated object detection. Covers the basic installation steps, a common CPU-fallback misconfiguration warning, and a specific PCIe `force_desc_page_size` fix needed to resolve a HailoRT runtime error. The Hailo-8L in the $70 AI HAT+ achieves ~12ms inference at low CPU usage and is noted as a cheaper alternative to the discontinued Google Coral TPU.

Frigate with Hailo for object detection on a Raspberry Pi