Frigate NVR Updates

Frigate NVR Updates: Weekly Recap - React 19 Upgrade and Motion Search Features

This week brought a major frontend upgrade to React 19 and introduced powerful new motion search capabilities. 23 pull requests were merged, focusing on UI improvements, debugging tools, and configuration enhancements.

Duration: PT2M49S

https://podlog.io/listen/frigate-nvr-updates-4774304c/episode/frigate-nvr-updates-weekly-recap-react-19-upgrade-and-motion-search-features-21677f69

Transcript

Good morning. This is your Frigate NVR weekly recap for March 1st through 8th, 2026.

23 PRs merged, 0 additional commits this week.

Starting with major features, the frontend received a significant upgrade to React 19, touching 50 files with over 4,500 line changes. This modernization includes new patches for Radix UI components and React WebSocket functionality.

Motion detection capabilities expanded substantially with the introduction of motion search functionality. This new feature adds API endpoints, background job processing, and database migrations to enable searching through motion events. The system now includes motion heatmap generation and improved review interfaces.

A notable addition for developers is the debug replay feature, allowing playback of recorded segments through the detection pipeline. This tool enables testing configuration changes against historical footage, streamlining the debugging process.

On the fixes front, several UI improvements landed. Browser zoom compatibility issues with mask and zone interfaces were resolved. The annotation offset adjustment interface received UX improvements, making it easier to calibrate object tracking overlays. Frontend compatibility issues following the React 19 upgrade were addressed through hook refactoring and state management updates.

Configuration management saw enhancements with dynamic field updates now supporting more parameters. Users can modify camera settings without full restarts in more scenarios. A new skip motion threshold option provides finer control over motion detection sensitivity.

Infrastructure improvements include better error handling in the Ollama GenAI integration and memory leak fixes in object description processing. License plate recognition logic was refined to apply filters after clustering, improving accuracy. Video playback reliability in tracking details was enhanced through better media handling.

Documentation updates clarified AVX2 CPU requirements across multiple features including face recognition, license plate detection, and semantic search. The getting started guide expanded the detectors section with additional configuration guidance.

Smaller but impactful changes include hiding alerts from dashboard-disabled cameras, adding fullscreen controls to tracking detail videos, and improved logging for anonymous user sessions. The region grid configuration now supports clearing operations directly from the frontend interface.

Development workflows benefited from updated GitHub Copilot instructions and improved test coverage for motion detection components.

Looking ahead, the React 19 foundation enables future frontend enhancements, while the new motion search infrastructure opens possibilities for advanced analytics features.

That's your Frigate NVR update for the week ending March 8th.