0.14.0 Beta 1

Beta Documentation: https://deploy-preview-11419--frigate-docs.netlify.app/

Major Changes for 0.14.0

Breaking Changes

There are several breaking changes in this release, so a backup of your current config and database is recommended before upgrading:

• Simply copy your current config file to a new location
• Stop Frigate and make a copy of the frigate.db file

Your existing events will not be visible in Frigate's new UI

• An entirely new way of reviewing camera footage has been implemented in Frigate 0.14, and existing events could not be migrated
• Recordings can still be viewed in the new UI using the recordings viewer (From Live view, click/tap on a camera and select the History button at the top left)
• You may want to export any starred/favorite events before upgrading
• If you are using the Home Assistant integration, events can still be viewed there

Zone/mask coordinates are now stored in the config as relative values

• Frigate's configuration migrator will read your existing mask/zone coordinates and convert them for relative use internally on startup
• After these changes, your config file will be incompatible with the prior version
• If you have mask/zone coordinate values that exceed the detect resolution, it will be ignored and you will see warnings in the log. Delete the mask or zone from your config and then use the editor in the UI to recreate it.

Backwards incompatible database changes

• There are a number of changes made to Frigate's database, which are not backwards compatible. A backup of the database is automatically made on startup, but a manual backup is highly recommended.

Brand New UI

Frigate 0.14's UI has been completely rewritten (guided demo) and optimized for desktop and mobile devices, aiming to simplify the user experience by focusing on two specific goals:

1. Allow you to quickly see what's currently happening or what just happened on your cameras. The new Live view features a real-time dashboard with live camera feeds and a filmstrip of animated thumbnails showing recent activity.
2. Make reviewing video footage more efficient. Frigate intelligently identifies parts of your timeline for review by leveraging the underlying data about motion and tracked objects. The new Review area implements a scrollable grid of thumbnails alongside a timeline where a hover or swipe begins a preview playback. Review items can be configured to be pre-categorized into alerts or detections based on cameras and/or zones. The review area also provides a motion review mode, allowing real-time scrubbing of video footage for any objects of interest that Frigate may have missed.

Frigate 0.14's UI will likely change the way you review your camera footage. It's a big change, but we love it and hope you do too. See the announcement post for more information regarding the motivation, ideas, and design decisions behind the new UI.

Dashboard

The new home page for Frigate is intended to be a real-time dashboard for your cameras. Recent alerts are represented by animated thumbnails in a filmstrip view above your camera feeds. Your camera images update once per minute when no detectable activity is occurring to conserve bandwidth and resources. As soon as any motion is detected, cameras seamlessly switch to a full-resolution live stream.

• Custom camera groups are now supported. Add any of your cameras (including the Birdseye camera) to user-definable groups and easily toggle between them.
• On desktops and tablets, cameras in a camera group can be resized and rearranged via dragging.

Live View

Clicking a camera from the dashboard will bring up a full-resolution live view where you can quickly access historical footage, use PTZ controls, toggle camera audio, use two-way talk, and more. Zooming with pinch or a scroll wheel is also supported.

Review

The Review area of Frigate 0.14 is for quickly reviewing historical footage of interest from your cameras. Review items are indicated on a vertical timeline and displayed as a grid of previews - bandwidth-optimized, low frame rate, low resolution videos. Hovering over or swiping a preview plays the video and marks it as reviewed. If more in-depth analysis is required, the preview can be clicked/tapped and the full frame rate, full resolution recording is displayed. Review items are filterable by date, object type, and camera.

Alerts and Detections

Not every segment of video captured by Frigate may be of the same level of interest to you. Video of people who enter your property are a different priority than those walking by on the sidewalk. For this reason, Frigate 0.14 categorizes review items as alerts and detections. By default, all person and car objects are considered alerts. You can refine categorization of your review items by configuring required zones for them.

See the review docs for more information

Motion review

In Frigate 0.13, there isn’t a way to see motion activity on a timeline and watch video at that time. The ability to scrub video for things Frigate missed has been one of the most common feature requests. Within the Motion tab of the Review section, you can now scrub through the timeline for all cameras simultaneously with a timeline showing relative motion activity. In order to avoid scrubbing over time periods already included in alerts and detections, you can also scrub in motion only mode.

History

In Frigate 0.14, full frame rate, full resolution camera footage (called "recordings" in earlier versions) can be viewed and scrubbed via a timeline. Previews of your other cameras are displayed and scrubbed alongside the main selected camera.

• Exporting footage from the History view has been drastically improved. Choose predefined time periods, enter a time directly, or drag handles on the timeline to select an export range.
• Past exports can be viewed, renamed, deleted, and downloaded from the dedicated Export pane, accessible from the navigation bar.

Other Notable UI Changes

• System stats - New charts and graphs with colored segments to track system usage history and more quickly identify problems or issues.
• Status bar - A new status bar gives at-a-glance CPU/GPU usage as well as any problems or issues with Frigate.
• New log viewer - The log viewer now provides a pane with links to solutions for common issues.
• Real-time motion tuning - Tune motion detection settings for your cameras without editing your config file and restarting.
• New mask and zone editor - Edit your masks and zones with a much-improved editor.

Authentication

Frigate now supports authentication with support for two different modes: native and proxy. In native mode, Frigate manages users and authentication using JWT tokens via a cookie or authorization header. Proxy mode is specifically designed to make use of common authentication proxies such as authelia, authentik, oauth2_proxy, etc.

• Authentication is enforced on a new port 8080, and port 5000 is still unauthenticated for backwards compatibility. If you want to use Frigate's authentication, update your reverse proxy to point at port 8080 instead of 5000.
• Support for role based authorization (admin/user roles) will come in a future release and will support upstream authentication proxies.
• We are planning to include TLS support in 0.14.0 as well.
• The Home Assistant integration needs to be pointed at port 5000. Some recommended configuration examples are in the updated docs.

See the documentation for more information.

Frigate+

Frigate 0.14 now offers a dedicated page for Frigate+ submissions, allowing more specific filtering by score and a faster workflow. Support has also been added to upload frames directly from the History view.

Other notable improvements

• Loitering Time - Zones now support a loitering_time configuration which can be used to only consider an object as part of a zone if they loiter for the specified number of seconds. This can be used, for example, to create alerts for cars that stop on the street but not cars that just drive past your camera.
• Two-way talk - For supported cameras, two-way talk can be used directly from a camera's live view.
• PWA-optimized - The UI has been optimized to work as a progressive web app. Pin Frigate to your iOS home screen or install on Android for a full app-like experience.
• Rockchip boards now support object detection - See the Rockchip documentation for more information.
• Automatic hardware acceleration - automatic detection and configuration of hardware acceleration for ffmpeg is attempted on startup
• New review related MQTT topics and API endpoints - New MQTT topics and api endpoints are available for reviews. These will likely be a better topic for notifications.

Fixes and chan...

0.13.2 Release

What's Changed

• Update Makefile for 0.13.2 by @NickM-27 in #9687
• Nms optimize for stationary cars by @NickM-27 in #9684
• More robust onvif handling by @hawkeye217 in #9635
• Don't fail if message is received before websocket start by @NickM-27 in #9634
• Birdseye enhancements by @NickM-27 in #9778
• Add detailed errors for unsupported label uploading to frigate+ by @Maximo-Guk in #9775
• Onvif: skip non-video profiles in setup by @hawkeye217 in #9708
• Mention that AMD CPUs work just fine with OpenVINO by @svenstaro in #9740
• Revise VSCode hostname info in docs by @printplaatreparatie in #9709
• Docs: Fix and clarify which /dev/video devices to use with Raspberry Pi by @m-d-brown in #9509
• remove reference to the term credit by @blakeblackshear in #9799
• Added documentation for time-lapse export (master) by @mike2307 in #9804
• docs: fix links to full configuration reference by @lawm in #9808
• Fix builds by @blakeblackshear in #9852
• Use zoom space id in Onvif RelativeMove setup by @hawkeye217 in #9859
• Fix birdseye camera comparison by @NickM-27 in #9887
• Format by @NickM-27 in #9889

New Contributors

• @Maximo-Guk made their first contribution in #9775
• @svenstaro made their first contribution in #9740
• @printplaatreparatie made their first contribution in #9709
• @m-d-brown made their first contribution in #9509
• @mike2307 made their first contribution in #9804
• @lawm made their first contribution in #9808

Full Changelog: v0.13.1...v0.13.2

0.13.1 Release

What's Changed

• Quick fix docs by @NickM-27 in #9506
• Increase hash map size by @blakeblackshear in #9515

Full Changelog: v0.13.0...v0.13.1

0.13.0 Release

Major Changes For 0.13.0

Security Advisories

The Github Security Lab team reached out to report the following security vulnerabilities.

• Cross-site request forgery in config_save and config_set request handlers (GHSL-2023-198)
• Reflected XSS through /<camera_name> API endpoints (GHSL-2023-195)
• Unsafe deserialization in load_config_with_no_duplicates of frigate/util/builtin.py (GHSL-2023-190)

Exploiting these vulnerabilities requires the attacker to both know very specific information about a user's Frigate server and requires an authenticated user to be tricked into clicking a specially crafted link to their Frigate instance.

This vulnerability could exploited by an attacker under the following circumstances:

1. Frigate publicly exposed to the internet (even with some authentication methods implemented)
2. Attacker knows the public address of a user's Frigate instance
3. Attacker crafts a specialized page which links to the user's Frigate instance
4. Attacker finds a way to get an authenticated user to visit their specialized page and click the button/link

If you haven't click any links to your Frigate server from unknown sources, there is no reason to believe that you have been compromised.

I have enabled the private vulnerability disclosure features inside Github for the project as well as the automated CodeQL scanning features to identify future issues.

Breaking Changes

• ARM 32-bit support has officially been removed. If you are running on a Raspberry Pi with a 32-bit OS, you will need to switch to 64-bit.
• You will likely experience some additional CPU usage because motion detection is now more sensitive and improve_contrast is now enabled by default. If you have specific values set in your config, it is recommended to remove them and re-calibrate as necessary.
• If no detect resolution is defined then frigate will now run detect at the native camera resolution. This means if you did not explicitly set the detect width / height AND your sub stream is not 1280x720, then your masks and zones will be configured incorrectly.
• Stationary object validation is now required, anyone who manually set interval: 0 will need to delete this field from the config or increase the value
• Changes have been made to the DB schema, downgrading to 0.12 will require restoring from a backup or deleting the DB and existing recordings / snapshots.
  NOTE: Some users have reported migrations taking a while and frigate frontend is not available until the migrations are complete.
• New location for frigate.db: Due to the support for network shares in HA OS and the popularity of storing recordings on a NAS in general, the database has a new default location of /config/frigate.db. This change is going to be done for existing users too, so frigate will automatically move the db for you.

  • For Frigate Addon Users:
    The migration is handled automatically and no action is required from the user.

  • For Docker Container Users:
    The migration is handled for you, but you must make sure that the docker CLI or docker-compose binds the entire /config directory and not just the /config/config.yml file.

    For example, if your current volume mapping is - /host/path/config_folder/config.yml:/config/config.yml it should now be - /host/path/config_folder/:/config/

• The previously deprecated record -> retain_days config has been removed, this should be migrated to record -> retain -> days if you have not already.
• HomeAssistant Integration 5.0 is required for the new features introduced in this release.
• TensorRT has been updated and the model generation has been refactored. New models will need to be generated for 0.13 and have a new path, check the detector docs for more info.
• TensorRT now requires Nvidia Driver Version 530+
• The cameras in /stats have been moved to their own cameras block. This only affects users that were directly reading from the /stats data via the API or MQTT.
• Removed direct access to the go2rtc http API to mitigate potential remote code execution via the exec source, you will need to map port 1984 and access directly if using go2rtc http api outside of frigate

Frigate+

Frigate+ Model Launch

Frigate+ models are now rolling out by invitation. If you would like to be invited for early access to purchase a subscription, please send an email to blake@frigate.video from the email address on your Frigate+ account. For more information, see this discussion.

False positive submissions

You can now submit false positive detections to Frigate+ for training feedback. Note that you will still need to login to your Frigate+ account and verify the images to ensure all relevant objects are labeled before the image will be used during training. Confirming the object is a true positive will also upload the image.

Models

This release has support for Frigate+ models. The model and related information are stored at /config/model_cache. Once Frigate+ models are launched, you will be able to use the model id in your config.

model:
  path: plus://your_model_id

Community Supported Boards

In version 0.12 Frigate gained support for Intel iGPUs via OpenVINO and for Nvidia GPUs, this was a great step forward for compatibility and for more users being able to use Frigate effectively. There are many other devices, including dedicated SBCs (Small Board Computers), that Frigate could support but the maintainers do not have the time or hardware to support.

The community supported boards framework will allow community members who want to contribute to create and manage support for a particular board. This will allow Frigate to support more devices while allowing the maintainers to focus on improving the core of Frigate.

Nvidia Jetson Support

@madsciencetist has contributed a community supported board build by adding support for Jetson 4.6 and Jetson 5 devices!

More information can be found in the detector docs on how to set this up.

https://docs.frigate.video/configuration/object_detectors#nvidia-tensorrt-detector

@MarcA711 has contributed a community supported board build by adding support for devices with RockChip SoCs:

RK3566/RK3568
RK3588/RK3588S
RV1103/RV1106
RK3562

https://docs.frigate.video/configuration/object_detectors#rockchip-rknn-toolkit-lite2

Object Tracking & Motion Detection

There have been many improvements to object tracking and motion detection which make Frigate more efficient and can reduce false positives.

New Object Tracker

Frigate now uses Norfair for object tracking, this means that object attributes like size, width/height ratio, and position are used to intelligently track objects. ID swapping between tracked objects is greatly reduced in situations where a car in the driveway was confused with a car passing in the street.

Improved Motion Detection

Frigate's motion detection is now more efficient. Motion detection will also re-calibrate for flashes like lightning or when the camera switches from color to IR mode or it is moved via PTZ. This reduces CPU usage during these changes and reduces false positives caused by excessive motion.

Historical Data

Frigate now uses the recordings timeline data to generate an 8x8 grid for each camera. This grid holds the range of expected regions (portions of the image sent to detection) for that area of the camera frame. This helps to improve accuracy of detection after motion has occurred and can reduce false positives in some cases.

The grid is updated on each startup and every day at 2 AM for each camera that is enabled.

Minimum Time For Zone Presence

Frigate now has a configurable field for the minimum amount of time an object needs to be in a zone before it is considered in a zone. For example, currently in Frigate if you have a front_yard zone and someone is walking by and briefly steps into the front yard Frigate would create an event for this. However, you may prefer for events to only be created if a person is in the front yard for multiple frames before creating an event.

zones:
  my_zone:
    inertia: 3  # <- 3 consecutive frames needed to consider object in zone
    coordinates: ...

Recordings Improvements

There have been a number of improvements to the management of recordings

Recordings Exporting

Frigate now supports exporting recordings in standard format or as a time-lapse. Exported recordings are stored in /media/frigate/exports and are downloadable from the UI.

Recordings Timeline Metadata

Frigate now saves metadata for key moments in an event such as when the tracked object is detected, enters or exits zone, becomes active / stationary, and leaves. This metadata is used to overlay on top of the recordings so it is easy to see what was detected and where. Note that the timing of the bounding boxes will likely be slightly off because it is based on the detect stream, but overlayed on the record stream. The annotation_offset value in the config can be used to adjust for any differences.

![Screen Shot 2023-07-27 ...

0.13.0 Release Candidate 1

Draft docs for this release: https://deploy-preview-6262--frigate-docs.netlify.app

Major Changes For 0.13.0

Security Advisories

The Github Security Lab team reached out to report the following security vulnerabilities.

• Cross-site request forgery in config_save and config_set request handlers (GHSL-2023-198)
• Reflected XSS through /<camera_name> API endpoints (GHSL-2023-195)
• Unsafe deserialization in load_config_with_no_duplicates of frigate/util/builtin.py (GHSL-2023-190)

Exploiting these vulnerabilities requires the attacker to both know very specific information about a user's Frigate server and requires an authenticated user to be tricked into clicking a specially crafted link to their Frigate instance.

This vulnerability could exploited by an attacker under the following circumstances:

1. Frigate publicly exposed to the internet (even with some authentication methods implemented)
2. Attacker knows the public address of a user's Frigate instance
3. Attacker crafts a specialized page which links to the user's Frigate instance
4. Attacker finds a way to get an authenticated user to visit their specialized page and click the button/link

If you haven't click any links to your Frigate server from unknown sources, there is no reason to believe that you have been compromised.

I have enabled the private vulnerability disclosure features inside Github for the project as well as the automated CodeQL scanning features to identify future issues.

Changes since Beta 7

• Compatibility for Jetpack 5.1.2 Jetson devices

Breaking Changes

• If no detect resolution is defined then frigate will now run detect at the native camera resolution. This means if you did not explicitly set the detect width / height AND your sub stream is not 1280x720, then your masks and zones will be configured incorrectly.
• Default values have been changed for motion detection and improve_contrast is now enabled by default. If you have specific values set in your config, it is recommended to remove them and re-calibrate as necessary.
• Stationary object validation is now required, anyone who manually set interval: 0 will need to delete this field from the config or increase the value
• Changes have been made to the DB schema, downgrading to 0.12 will require restoring from a backup or deleting the DB and existing recordings / snapshots.
  NOTE: Some users have reported migrations taking a while and frigate frontend is not available until the migrations are complete.
• New location for frigate.db: Due to the support for network shares in HA OS and the popularity of storing recordings on a NAS in general, the database has a new default location of /config/frigate.db. This change is going to be done for existing users too, so frigate will automatically move the db for you.

  • For Frigate Addon Users:
    The migration is handled automatically and no action is required from the user.

  • For Docker Container Users:
    The migration is handled for you, but you must make sure that the docker CLI or docker-compose binds the entire /config directory and not just the /config/config.yml file.

    For example, if your current volume mapping is - /host/path/config_folder/config.yml:/config/config.yml it should now be - /host/path/config_folder/:/config/

• The previously deprecated record -> retain_days config has been removed, this should be migrated to record -> retain -> days if you have not already.
• HomeAssistant Integration 5.0 beta is required for the new features introduced in this release.
• TensorRT has been updated and the model generation has been refactored. New models will need to be generated for 0.13 and have a new path, check the detector docs for more info.
• TensorRT now requires Nvidia Driver Version 530+
• The cameras in /stats have been moved to their own cameras block. This only affects users that were directly reading from the /stats data via the API or MQTT.

Frigate+

Frigate+ Model Launch

Frigate+ models are now rolling out by invitation. If you would like to be invited for early access to purchase a subscription, please send an email to blake@frigate.video from the email address on your Frigate+ account. For more information, see this discussion.

False positive submissions

You can now submit false positive detections to Frigate+ for training feedback. Note that you will still need to login to your Frigate+ account and verify the images to ensure all relevant objects are labeled before the image will be used during training. Confirming the object is a true positive will also upload the image.

Models

This release has support for Frigate+ models. The model and related information are stored at /config/model_cache. Once Frigate+ models are launched, you will be able to use the model id in your config.

model:
  path: plus://your_model_id

Community Supported Boards

In version 0.12 Frigate gained support for Intel iGPUs via OpenVINO and for Nvidia GPUs, this was a great step forward for compatibility and for more users being able to use Frigate effectively. There are many other devices, including dedicated SBCs (Small Board Computers), that Frigate could support but the maintainers do not have the time or hardware to support.

The community supported boards framework will allow community members who want to contribute to create and manage support for a particular board. This will allow Frigate to support more devices while allowing the maintainers to focus on improving the core of Frigate.

Nvidia Jetson Support

@madsciencetist has contributed a community supported board build by adding support for Jetson 4.6 and Jetson 5 devices!

More information can be found in the detector docs on how to set this up.

https://deploy-preview-6262--frigate-docs.netlify.app/configuration/object_detectors#nvidia-tensorrt-detector

@MarcA711 has contributed a community supported board build by adding support for devices with RockChip SoCs:

RK3566/RK3568
RK3588/RK3588S
RV1103/RV1106
RK3562

https://deploy-preview-6262--frigate-docs.netlify.app/configuration/object_detectors#rockchip-rknn-toolkit-lite2

Object Tracking & Motion Detection

There have been many improvements to object tracking and motion detection which make Frigate more efficient and can reduce false positives.

New Object Tracker

Frigate now uses Norfair for object tracking, this means that object attributes like size, width/height ratio, and position are used to intelligently track objects. ID swapping between tracked objects is greatly reduced in situations where a car in the driveway was confused with a car passing in the street.

Improved Motion Detection

Frigate's motion detection is now more efficient. Motion detection will also re-calibrate for flashes like lightning or when the camera switches from color to IR mode or it is moved via PTZ. This reduces CPU usage during these changes and reduces false positives caused by excessive motion.

Historical Data

Frigate now uses the recordings timeline data to generate an 8x8 grid for each camera. This grid holds the range of expected regions (portions of the image sent to detection) for that area of the camera frame. This helps to improve accuracy of detection after motion has occurred and can reduce false positives in some cases.

The grid is updated on each startup and every day at 2 AM for each camera that is enabled.

Minimum Time For Zone Presence

Frigate now has a configurable field for the minimum amount of time an object needs to be in a zone before it is considered in a zone. For example, currently in Frigate if you have a front_yard zone and someone is walking by and briefly steps into the front yard Frigate would create an event for this. However, you may prefer for events to only be created if a person is in the front yard for multiple frames before creating an event.

zones:
  my_zone:
    inertia: 3  # <- 3 consecutive frames needed to consider object in zone
    coordinates: ...

Recordings Improvements

There have been a number of improvements to the management of recordings

Recordings Exporting

Frigate now supports exporting recordings in standard format or as a time-lapse. Exported recordings are stored in /media/frigate/exports and are downloadable from the UI.

Recordings Timeline Metadata

Frigate now saves metadata for key moments in an event such as when the tracked object is detected, enters or exits zone, becomes active / stationary, and leaves. This metadata is used to overlay on top of the recordings so it is easy to see what was detected and where. Note that the timing of the bounding boxes will likely be slightly off because it is based on the detect stream, but overlayed on the record stream. The annotation_offset value in the config can be used to adjust for any differences.

Internal Processing...

0.13.0 Beta 7

Beta docs for this release: https://deploy-preview-6262--frigate-docs.netlify.app

Major Changes For 0.13.0

Security Advisories

The Github Security Lab team reached out to report the following security vulnerabilities.

• Cross-site request forgery in config_save and config_set request handlers (GHSL-2023-198)
• Reflected XSS through /<camera_name> API endpoints (GHSL-2023-195)
• Unsafe deserialization in load_config_with_no_duplicates of frigate/util/builtin.py (GHSL-2023-190)

Exploiting these vulnerabilities requires the attacker to both know very specific information about a user's Frigate server and requires an authenticated user to be tricked into clicking a specially crafted link to their Frigate instance.

This vulnerability could exploited by an attacker under the following circumstances:

1. Frigate publicly exposed to the internet (even with some authentication methods implemented)
2. Attacker knows the public address of a user's Frigate instance
3. Attacker crafts a specialized page which links to the user's Frigate instance
4. Attacker finds a way to get an authenticated user to visit their specialized page and click the button/link

If you haven't click any links to your Frigate server from unknown sources, there is no reason to believe that you have been compromised.

I have enabled the private vulnerability disclosure features inside Github for the project as well as the automated CodeQL scanning features to identify future issues.

Changes since Beta 6

• Remove direct access to the go2rtc API to mitigate potential remote code execution via the exec source
• Other small fixes and improvements

Breaking Changes

• If no detect resolution is defined then frigate will now run detect at the native camera resolution. This means if you did not explicitly set the detect width / height AND your sub stream is not 1280x720, then your masks and zones will be configured incorrectly.
• Default values have been changed for motion detection and improve_contrast is now enabled by default. If you have specific values set in your config, it is recommended to remove them and re-calibrate as necessary.
• Stationary object validation is now required, anyone who manually set interval: 0 will need to delete this field from the config or increase the value
• Changes have been made to the DB schema, downgrading to 0.12 will require restoring from a backup or deleting the DB and existing recordings / snapshots.
  NOTE: Some users have reported migrations taking a while and frigate frontend is not available until the migrations are complete.
• New location for frigate.db: Due to the support for network shares in HA OS and the popularity of storing recordings on a NAS in general, the database has a new default location of /config/frigate.db. This change is going to be done for existing users too, so frigate will automatically move the db for you.

  • For Frigate Addon Users:
    The migration is handled automatically and no action is required from the user.

  • For Docker Container Users:
    The migration is handled for you, but you must make sure that the docker CLI or docker-compose binds the entire /config directory and not just the /config/config.yml file.

    For example, if your current volume mapping is - /host/path/config_folder/config.yml:/config/config.yml it should now be - /host/path/config_folder/:/config/

• The previously deprecated record -> retain_days config has been removed, this should be migrated to record -> retain -> days if you have not already.
• HomeAssistant Integration 5.0 beta is required for the new features introduced in this release.
• TensorRT has been updated and the model generation has been refactored. New models will need to be generated for 0.13 and have a new path, check the detector docs for more info.
• TensorRT now requires Nvidia Driver Version 530+
• The cameras in /stats have been moved to their own cameras block. This only affects users that were directly reading from the /stats data via the API or MQTT.

Frigate+

Frigate+ Model Launch

Frigate+ models are now rolling out by invitation. If you would like to be invited for early access to purchase a subscription, please send an email to blake@frigate.video from the email address on your Frigate+ account. For more information, see this discussion.

False positive submissions

You can now submit false positive detections to Frigate+ for training feedback. Note that you will still need to login to your Frigate+ account and verify the images to ensure all relevant objects are labeled before the image will be used during training. Confirming the object is a true positive will also upload the image.

Models

Frigate+ models are almost ready for launch, and this release has support for automatically downloading from Frigate+. The model and related information are stored at /config/model_cache. Once Frigate+ models are launched, you will be able to use the model id in your config.

model:
  path: plus://e63b7345cc83a84ed79dedfc99c16616

Community Supported Boards

In version 0.12 Frigate gained support for Intel iGPUs via OpenVINO and for Nvidia GPUs, this was a great step forward for compatibility and for more users being able to use Frigate effectively. There are many other devices, including dedicated SBCs (Small Board Computers), that Frigate could support but the maintainers do not have the time or hardware to support.

The community supported boards framework will allow community members who want to contribute to create and manage support for a particular board. This will allow Frigate to support more devices while allowing the maintainers to focus on improving the core of Frigate.

Nvidia Jetson Support

@madsciencetist has contributed a community supported board build by adding support for Jetson 4.6 and Jetson 5 devices!

More information can be found in the detector docs on how to set this up.

https://deploy-preview-6262--frigate-docs.netlify.app/configuration/object_detectors#nvidia-tensorrt-detector

@MarcA711 has contributed a community supported board build by adding support for devices with RockChip SoCs:

RK3566/RK3568
RK3588/RK3588S
RV1103/RV1106
RK3562

https://deploy-preview-6262--frigate-docs.netlify.app/configuration/object_detectors#rockchip-rknn-toolkit-lite2

Object Tracking & Motion Detection

There have been many improvements to object tracking and motion detection which make Frigate more efficient and can reduce false positives.

New Object Tracker

Frigate now uses Norfair for object tracking, this means that object attributes like size, width/height ratio, and position are used to intelligently track objects. ID swapping between tracked objects is greatly reduced in situations where a car in the driveway was confused with a car passing in the street.

Improved Motion Detection

Frigate's motion detection is now more efficient. Motion detection will also re-calibrate for flashes like lightning or when the camera switches from color to IR mode or it is moved via PTZ. This reduces CPU usage during these changes and reduces false positives caused by excessive motion.

Historical Data

Frigate now uses the recordings timeline data to generate an 8x8 grid for each camera. This grid holds the range of expected regions (portions of the image sent to detection) for that area of the camera frame. This helps to improve accuracy of detection after motion has occurred and can reduce false positives in some cases.

The grid is updated on each startup and every day at 2 AM for each camera that is enabled.

Minimum Time For Zone Presence

Frigate now has a configurable field for the minimum amount of time an object needs to be in a zone before it is considered in a zone. For example, currently in Frigate if you have a front_yard zone and someone is walking by and briefly steps into the front yard Frigate would create an event for this. However, you may prefer for events to only be created if a person is in the front yard for multiple frames before creating an event.

zones:
  my_zone:
    inertia: 3  # <- 3 consecutive frames needed to consider object in zone
    coordinates: ...

Recordings Improvements

There have been a number of improvements to the management of recordings

Recordings Exporting

Frigate now supports exporting recordings in standard format or as a time-lapse. Exported recordings are stored in /media/frigate/exports and are downloadable from the UI.

Recordings Timeline Metadata

Frigate now saves metadata for key moments in an event such as when the tracked object is detected, enters or exits zone, becomes active / stationary, and leaves. This metadata is used to overlay on top of the recordings so it is easy to see what was detected and where. Note that the timing of the bounding boxes will likely be slightly off because it is based on the detect stream, but overlayed on the record stream. The annotation_offset value in the config can be used to adjust for...

0.13.0 Beta 6

Beta docs for this release: https://deploy-preview-6262--frigate-docs.netlify.app

Major Changes For 0.13.0

Security Advisories

The Github Security Lab team reached out to report the following security vulnerabilities.

• Cross-site request forgery in config_save and config_set request handlers (GHSL-2023-198)
• Reflected XSS through /<camera_name> API endpoints (GHSL-2023-195)
• Unsafe deserialization in load_config_with_no_duplicates of frigate/util/builtin.py (GHSL-2023-190)

Exploiting these vulnerabilities requires the attacker to both know very specific information about a user's Frigate server and requires an authenticated user to be tricked into clicking a specially crafted link to their Frigate instance.

This vulnerability could exploited by an attacker under the following circumstances:

1. Frigate publicly exposed to the internet (even with some authentication methods implemented)
2. Attacker knows the public address of a user's Frigate instance
3. Attacker crafts a specialized page which links to the user's Frigate instance
4. Attacker finds a way to get an authenticated user to visit their specialized page and click the button/link

If you haven't click any links to your Frigate server from unknown sources, there is no reason to believe that you have been compromised.

I have enabled the private vulnerability disclosure features inside Github for the project as well as the automated CodeQL scanning features to identify future issues.

Changes since Beta 5

• The option record -> sync_on_startup has been renamed record -> sync_recordings which controls startup and periodic recordings. This is a breaking change only for beta users who have enabled this option.
• community supported build added for certain rockchip SOCs
• Go2rtc updated to 1.8.4

Breaking Changes

• If no detect resolution is defined then frigate will now run detect at the native camera resolution. This means if you did not explicitly set the detect width / height AND your sub stream is not 1280x720, then your masks and zones will be configured incorrectly.
• Default values have been changed for motion detection and improve_contrast is now enabled by default. If you have specific values set in your config, it is recommended to remove them and re-calibrate as necessary.
• Stationary object validation is now required, anyone who manually set interval: 0 will need to delete this field from the config or increase the value
• Changes have been made to the DB schema, downgrading to 0.12 will require restoring from a backup or deleting the DB and existing recordings / snapshots.
  NOTE: Some users have reported migrations taking a while and frigate frontend is not available until the migrations are complete.
• New location for frigate.db: Due to the support for network shares in HA OS and the popularity of storing recordings on a NAS in general, the database has a new default location of /config/frigate.db. This change is going to be done for existing users too, so frigate will automatically move the db for you.

  • For Frigate Addon Users:
    The migration is handled automatically and no action is required from the user.

  • For Docker Container Users:
    The migration is handled for you, but you must make sure that the docker CLI or docker-compose binds the entire /config directory and not just the /config/config.yml file.

    For example, if your current volume mapping is - /host/path/config_folder/config.yml:/config/config.yml it should now be - /host/path/config_folder/:/config/

• The previously deprecated record -> retain_days config has been removed, this should be migrated to record -> retain -> days if you have not already.
• HomeAssistant Integration 5.0 beta is required for the new features introduced in this release.
• TensorRT has been updated and the model generation has been refactored. New models will need to be generated for 0.13 and have a new path, check the detector docs for more info.
• TensorRT now requires Nvidia Driver Version 530+
• The cameras in /stats have been moved to their own cameras block. This only affects users that were directly reading from the /stats data via the API or MQTT.

Frigate+

Frigate+ Model Launch

Frigate+ models are now rolling out by invitation. If you would like to be invited for early access to purchase a subscription, please send an email to blake@frigate.video from the email address on your Frigate+ account. For more information, see this discussion.

False positive submissions

You can now submit false positive detections to Frigate+ for training feedback. Note that you will still need to login to your Frigate+ account and verify the images to ensure all relevant objects are labeled before the image will be used during training. Confirming the object is a true positive will also upload the image.

Models

Frigate+ models are almost ready for launch, and this release has support for automatically downloading from Frigate+. The model and related information are stored at /config/model_cache. Once Frigate+ models are launched, you will be able to use the model id in your config.

model:
  path: plus://e63b7345cc83a84ed79dedfc99c16616

Community Supported Boards

In version 0.12 Frigate gained support for Intel iGPUs via OpenVINO and for Nvidia GPUs, this was a great step forward for compatibility and for more users being able to use Frigate effectively. There are many other devices, including dedicated SBCs (Small Board Computers), that Frigate could support but the maintainers do not have the time or hardware to support.

The community supported boards framework will allow community members who want to contribute to create and manage support for a particular board. This will allow Frigate to support more devices while allowing the maintainers to focus on improving the core of Frigate.

Nvidia Jetson Support

@madsciencetist has contributed a community supported board build by adding support for Jetson 4.6 and Jetson 5 devices!

More information can be found in the detector docs on how to set this up.

https://deploy-preview-6262--frigate-docs.netlify.app/configuration/object_detectors#nvidia-tensorrt-detector

@MarcA711 has contributed a community supported board build by adding support for devices with RockChip SoCs:

RK3566/RK3568
RK3588/RK3588S
RV1103/RV1106
RK3562

https://deploy-preview-6262--frigate-docs.netlify.app/configuration/object_detectors#rockchip-rknn-toolkit-lite2

Object Tracking & Motion Detection

There have been many improvements to object tracking and motion detection which make Frigate more efficient and can reduce false positives.

New Object Tracker

Frigate now uses Norfair for object tracking, this means that object attributes like size, width/height ratio, and position are used to intelligently track objects. ID swapping between tracked objects is greatly reduced in situations where a car in the driveway was confused with a car passing in the street.

Improved Motion Detection

Frigate's motion detection is now more efficient. Motion detection will also re-calibrate for flashes like lightning or when the camera switches from color to IR mode or it is moved via PTZ. This reduces CPU usage during these changes and reduces false positives caused by excessive motion.

Historical Data

Frigate now uses the recordings timeline data to generate an 8x8 grid for each camera. This grid holds the range of expected regions (portions of the image sent to detection) for that area of the camera frame. This helps to improve accuracy of detection after motion has occurred and can reduce false positives in some cases.

The grid is updated on each startup and every day at 2 AM for each camera that is enabled.

Minimum Time For Zone Presence

Frigate now has a configurable field for the minimum amount of time an object needs to be in a zone before it is considered in a zone. For example, currently in Frigate if you have a front_yard zone and someone is walking by and briefly steps into the front yard Frigate would create an event for this. However, you may prefer for events to only be created if a person is in the front yard for multiple frames before creating an event.

zones:
  my_zone:
    inertia: 3  # <- 3 consecutive frames needed to consider object in zone
    coordinates: ...

Recordings Improvements

There have been a number of improvements to the management of recordings

Recordings Exporting

Frigate now supports exporting recordings in standard format or as a time-lapse. Exported recordings are stored in /media/frigate/exports and are downloadable from the UI.

Recordings Timeline Metadata

Frigate now saves metadata for key moments in an event such as when the tracked object is detected, enters or exits zone, becomes active / stationary, and leaves. This metadata is used to overlay on top of the recordings so it is easy to see what was detected and where. Note that the timing of the bounding boxes will likely be sligh...

0.13.0 Beta 5

Beta docs for this release: https://deploy-preview-6262--frigate-docs.netlify.app

Major Changes For 0.13.0

Security Advisories

The Github Security Lab team reached out to report the following security vulnerabilities.

• Cross-site request forgery in config_save and config_set request handlers (GHSL-2023-198)
• Reflected XSS through /<camera_name> API endpoints (GHSL-2023-195)
• Unsafe deserialization in load_config_with_no_duplicates of frigate/util/builtin.py (GHSL-2023-190)

Exploiting these vulnerabilities requires the attacker to both know very specific information about a user's Frigate server and requires an authenticated user to be tricked into clicking a specially crafted link to their Frigate instance.

This vulnerability could exploited by an attacker under the following circumstances:

1. Frigate publicly exposed to the internet (even with some authentication methods implemented)
2. Attacker knows the public address of a user's Frigate instance
3. Attacker crafts a specialized page which links to the user's Frigate instance
4. Attacker finds a way to get an authenticated user to visit their specialized page and click the button/link

If you haven't click any links to your Frigate server from unknown sources, there is no reason to believe that you have been compromised.

I have enabled the private vulnerability disclosure features inside Github for the project as well as the automated CodeQL scanning features to identify future issues.

Changes since Beta 4

• NOTE: An issue was discovered where timeline entries were saved in the DB without a corresponding event. This has been fixed and the DB will be retroactively cleaned up upon first starting beta5. Please note that frigate may be slow to start the first time but this will also reduce the DB size considerably.
• Improve recordings database and fix writing recordings without timezone info (important for DST)
• Update go2rtc to 1.8.2 and fix go2rtc default webrtc config
• Autotracking improvements
• Improve handling of output so stuck websockets don't cause frame processing to backup
• Only cache json requests with nginx

Breaking Changes

• If no detect resolution is defined then frigate will now run detect at the native camera resolution. This means if you did not explicitly set the detect width / height AND your sub stream is not 1280x720, then your masks and zones will be configured incorrectly.
• Default values have been changed for motion detection. If you have specific values set in your config, it is recommended to remove them and re-calibrate as necessary.
• Stationary object validation is now required, anyone who manually set interval: 0 will need to delete this field from the config or increase the value
• Changes have been made to the DB schema, downgrading to 0.12 will require restoring from a backup or deleting the DB and existing recordings / snapshots.
  NOTE: Some users have reported migrations taking a while and frigate frontend is not available until the migrations are complete.
• New location for frigate.db: Due to the support for network shares in HA OS and the popularity of storing recordings on a NAS in general, the database has a new default location of /config/frigate.db. This change is going to be done for existing users too, so frigate will automatically move the db for you.

  • For Frigate Addon Users:
    The migration is handled automatically and no action is required from the user.

  • For Docker Container Users:
    The migration is handled for you, but you must make sure that the docker CLI or docker-compose binds the entire /config directory and not just the /config/config.yml file.

    For example, if your current volume mapping is - /host/path/config_folder/config.yml:/config/config.yml it should now be - /host/path/config_folder/:/config/

• The previously deprecated record -> retain_days config has been removed, this should be migrated to record -> retain -> days if you have not already.
• HomeAssistant Integration 5.0 beta is required for the new features introduced in this release.
• TensorRT has been updated and the model generation has been refactored. New models will need to be generated for 0.13 and have a new path, check the detector docs for more info.
• TensorRT now requires Nvidia Driver Version 530+
• The cameras in /stats have been moved to their own cameras block. This only affects users that were directly reading from the /stats data via the API or MQTT.

Frigate+

Frigate+ Model Launch

Frigate+ models are now rolling out by invitation. If you would like to be invited for early access to purchase a subscription, please send an email to blake@frigate.video from the email address on your Frigate+ account. For more information, see this discussion.

False positive submissions

You can now submit false positive detections to Frigate+ for training feedback. Note that you will still need to login to your Frigate+ account and verify the images to ensure all relevant objects are labeled before the image will be used during training. Confirming the object is a true positive will also upload the image.

Models

Frigate+ models are almost ready for launch, and this release has support for automatically downloading from Frigate+. The model and related information are stored at /config/model_cache. Once Frigate+ models are launched, you will be able to use the model id in your config.

model:
  path: plus://e63b7345cc83a84ed79dedfc99c16616

Community Supported Boards

In version 0.12 Frigate gained support for Intel iGPUs via OpenVINO and for Nvidia GPUs, this was a great step forward for compatibility and for more users being able to use Frigate effectively. There are many other devices, including dedicated SBCs (Small Board Computers), that Frigate could support but the maintainers do not have the time or hardware to support.

The community supported boards framework will allow community members who want to contribute to create and manage support for a particular board. This will allow Frigate to support more devices while allowing the maintainers to focus on improving the core of Frigate.

Nvidia Jetson Support

@madsciencetist has contributed a community supported board build by adding support for Jetson 4.6 and Jetson 5 devices!

More information can be found in the detector docs on how to set this up.

https://deploy-preview-6262--frigate-docs.netlify.app/configuration/object_detectors#nvidia-tensorrt-detector

@MarcA711 has contributed a community supported board build by adding support for devices with RockChip SoCs:

RK3566/RK3568
RK3588/RK3588S
RV1103/RV1106
RK3562

https://deploy-preview-6262--frigate-docs.netlify.app/configuration/object_detectors#rockchip-rknn-toolkit-lite2

Object Tracking & Motion Detection

There have been many improvements to object tracking and motion detection which make Frigate more efficient and can reduce false positives.

New Object Tracker

Frigate now uses Norfair for object tracking, this means that object attributes like size, width/height ratio, and position are used to intelligently track objects. ID swapping between tracked objects is greatly reduced in situations where a car in the driveway was confused with a car passing in the street.

Improved Motion Detection

Frigate's motion detection is now more efficient. Motion detection will also re-calibrate for flashes like lightning or when the camera switches from color to IR mode or it is moved via PTZ. This reduces CPU usage during these changes and reduces false positives caused by excessive motion.

Historical Data

Frigate now uses the recordings timeline data to generate an 8x8 grid for each camera. This grid holds the range of expected regions (portions of the image sent to detection) for that area of the camera frame. This helps to improve accuracy of detection after motion has occurred and can reduce false positives in some cases.

The grid is updated on each startup and every day at 2 AM for each camera that is enabled.

Minimum Time For Zone Presence

Frigate now has a configurable field for the minimum amount of time an object needs to be in a zone before it is considered in a zone. For example, currently in Frigate if you have a front_yard zone and someone is walking by and briefly steps into the front yard Frigate would create an event for this. However, you may prefer for events to only be created if a person is in the front yard for multiple frames before creating an event.

zones:
  my_zone:
    inertia: 3  # <- 3 consecutive frames needed to consider object in zone
    coordinates: ...

Recordings Improvements

There have been a number of improvements to the management of recordings

Recordings Exporting

Frigate now supports exporting recordings in standard format or as a time-lapse. Exported recordings are stored in /media/frigate/exports and are downloadable from the UI.

Recordings Timeline Metadata

Frigate now saves metadata for key moments in ...

0.13.0 Beta 4

Beta docs for this release: https://deploy-preview-6262--frigate-docs.netlify.app

Major Changes For 0.13.0

Security Advisories

The Github Security Lab team reached out to report the following security vulnerabilities.

• Cross-site request forgery in config_save and config_set request handlers (GHSL-2023-198)
• Reflected XSS through /<camera_name> API endpoints (GHSL-2023-195)
• Unsafe deserialization in load_config_with_no_duplicates of frigate/util/builtin.py (GHSL-2023-190)

Exploiting these vulnerabilities requires the attacker to both know very specific information about a user's Frigate server and requires an authenticated user to be tricked into clicking a specially crafted link to their Frigate instance.

This vulnerability could exploited by an attacker under the following circumstances:

1. Frigate publicly exposed to the internet (even with some authentication methods implemented)
2. Attacker knows the public address of a user's Frigate instance
3. Attacker crafts a specialized page which links to the user's Frigate instance
4. Attacker finds a way to get an authenticated user to visit their specialized page and click the button/link

If you haven't click any links to your Frigate server from unknown sources, there is no reason to believe that you have been compromised.

I have enabled the private vulnerability disclosure features inside Github for the project as well as the automated CodeQL scanning features to identify future issues.

Changes since Beta 3

• Nginx caching to better handle rapid fire API requests from third party requests
• Fix daylight savings time issue with recordings validation
• More improvements to birdseye logic
• Devices with RockChip SoC are now supported as a community build. See docs for more info.

Breaking Changes

• If no detect resolution is defined then frigate will now run detect at the native camera resolution. This means if you did not explicitly set the detect width / height AND your sub stream is not 1280x720, then your masks and zones will be configured incorrectly.
• Default values have been changed for motion detection. If you have specific values set in your config, it is recommended to remove them and re-calibrate as necessary.
• Stationary object validation is now required, anyone who manually set interval: 0 will need to delete this field from the config or increase the value
• Changes have been made to the DB schema, downgrading to 0.12 will require restoring from a backup or deleting the DB and existing recordings / snapshots.
  NOTE: Some users have reported migrations taking a while and frigate frontend is not available until the migrations are complete.
• New location for frigate.db: Due to the support for network shares in HA OS and the popularity of storing recordings on a NAS in general, the database has a new default location of /config/frigate.db. This change is going to be done for existing users too, so frigate will automatically move the db for you.

  • For Frigate Addon Users:
    The migration is handled automatically and no action is required from the user.

  • For Docker Container Users:
    The migration is handled for you, but you must make sure that the docker CLI or docker-compose binds the entire /config directory and not just the /config/config.yml file.

    For example, if your current volume mapping is - /host/path/config_folder/config.yml:/config/config.yml it should now be - /host/path/config_folder/:/config/

• The previously deprecated record -> retain_days config has been removed, this should be migrated to record -> retain -> days if you have not already.
• HomeAssistant Integration 5.0 beta is required for the new features introduced in this release.
• TensorRT has been updated and the model generation has been refactored. New models will need to be generated for 0.13 and have a new path, check the detector docs for more info.
• TensorRT now requires Nvidia Driver Version 530+
• The cameras in /stats have been moved to their own cameras block. This only affects users that were directly reading from the /stats data via the API or MQTT.

Frigate+

Frigate+ Model Launch

Frigate+ models are now rolling out by invitation. If you would like to be invited for early access to purchase a subscription, please send an email to blake@frigate.video from the email address on your Frigate+ account. For more information, see this discussion.

False positive submissions

You can now submit false positive detections to Frigate+ for training feedback. Note that you will still need to login to your Frigate+ account and verify the images to ensure all relevant objects are labeled before the image will be used during training. Confirming the object is a true positive will also upload the image.

Models

Frigate+ models are almost ready for launch, and this release has support for automatically downloading from Frigate+. The model and related information are stored at /config/model_cache. Once Frigate+ models are launched, you will be able to use the model id in your config.

model:
  path: plus://e63b7345cc83a84ed79dedfc99c16616

Community Supported Boards

In version 0.12 Frigate gained support for Intel iGPUs via OpenVINO and for Nvidia GPUs, this was a great step forward for compatibility and for more users being able to use Frigate effectively. There are many other devices, including dedicated SBCs (Small Board Computers), that Frigate could support but the maintainers do not have the time or hardware to support.

The community supported boards framework will allow community members who want to contribute to create and manage support for a particular board. This will allow Frigate to support more devices while allowing the maintainers to focus on improving the core of Frigate.

Nvidia Jetson Support

@madsciencetist has contributed a community supported board build by adding support for Jetson 4.6 and Jetson 5 devices!

More information can be found in the detector docs on how to set this up.

https://deploy-preview-6262--frigate-docs.netlify.app/configuration/object_detectors#nvidia-tensorrt-detector

@MarcA711 has contributed a community supported board build by adding support for devices with RockChip SoCs:

RK3566/RK3568
RK3588/RK3588S
RV1103/RV1106
RK3562

https://deploy-preview-6262--frigate-docs.netlify.app/configuration/object_detectors#rockchip-rknn-toolkit-lite2

Object Tracking & Motion Detection

There have been many improvements to object tracking and motion detection which make Frigate more efficient and can reduce false positives.

New Object Tracker

Frigate now uses Norfair for object tracking, this means that object attributes like size, width/height ratio, and position are used to intelligently track objects. ID swapping between tracked objects is greatly reduced in situations where a car in the driveway was confused with a car passing in the street.

Improved Motion Detection

Frigate's motion detection is now more efficient. Motion detection will also re-calibrate for flashes like lightning or when the camera switches from color to IR mode or it is moved via PTZ. This reduces CPU usage during these changes and reduces false positives caused by excessive motion.

Historical Data

Frigate now uses the recordings timeline data to generate an 8x8 grid for each camera. This grid holds the range of expected regions (portions of the image sent to detection) for that area of the camera frame. This helps to improve accuracy of detection after motion has occurred and can reduce false positives in some cases.

The grid is updated on each startup and every day at 2 AM for each camera that is enabled.

Minimum Time For Zone Presence

Frigate now has a configurable field for the minimum amount of time an object needs to be in a zone before it is considered in a zone. For example, currently in Frigate if you have a front_yard zone and someone is walking by and briefly steps into the front yard Frigate would create an event for this. However, you may prefer for events to only be created if a person is in the front yard for multiple frames before creating an event.

zones:
  my_zone:
    inertia: 3  # <- 3 consecutive frames needed to consider object in zone
    coordinates: ...

Recordings Improvements

There have been a number of improvements to the management of recordings

Recordings Exporting

Frigate now supports exporting recordings in standard format or as a time-lapse. Exported recordings are stored in /media/frigate/exports and are downloadable from the UI.

Recordings Timeline Metadata

Frigate now saves metadata for key moments in an event such as when the tracked object is detected, enters or exits zone, becomes active / stationary, and leaves. This metadata is used to overlay on top of the recordings so it is easy to see what was detected and where. Note that the timing of the bounding boxes will likely be slightly off because it is based on the detect stream, but overlayed on the record stream. The annotation_offset value in the config can be used to adjust for an...

0.13.0 Beta 3

Beta docs for this release: https://deploy-preview-6262--frigate-docs.netlify.app

Major Changes For 0.13.0

Security Advisories

The Github Security Lab team reached out to report the following security vulnerabilities.

• Cross-site request forgery in config_save and config_set request handlers (GHSL-2023-198)
• Reflected XSS through /<camera_name> API endpoints (GHSL-2023-195)
• Unsafe deserialization in load_config_with_no_duplicates of frigate/util/builtin.py (GHSL-2023-190)

Exploiting these vulnerabilities requires the attacker to both know very specific information about a user's Frigate server and requires an authenticated user to be tricked into clicking a specially crafted link to their Frigate instance.

This vulnerability could exploited by an attacker under the following circumstances:

1. Frigate publicly exposed to the internet (even with some authentication methods implemented)
2. Attacker knows the public address of a user's Frigate instance
3. Attacker crafts a specialized page which links to the user's Frigate instance
4. Attacker finds a way to get an authenticated user to visit their specialized page and click the button/link

If you haven't click any links to your Frigate server from unknown sources, there is no reason to believe that you have been compromised.

I have enabled the private vulnerability disclosure features inside Github for the project as well as the automated CodeQL scanning features to identify future issues.

Changes since Beta 2

• A breaking change to the integration was introduced. You will need to update to the beta version of the integration.
• Frigate will now use the historical entries in the recordings timeline to create a grid of the expected size of objects, this will improve the accuracy of detection when motion is first detected. Motion is also optimized to reduce cases where a persons legs or arms were detected separately which led to multiple events for the same object.
• Many optimizations around stationary objects to optimize frequency and improve accuracy of checks for stationary objects.
• Motion detection calibration has been revamped to be dynamic, this is important for cases where motion was occurring after a large change; for example when someone is walking in the frame after moving a ptz camera.
• New recording timeline entries for Frigate+ attributes that show the moment an attribute (amazon, face, fedex, license_plate, ups) were first detected.
• Many camera auto tracking improvements including optimizations to movement, zoom, and camera calibration.
• Ongoing and past events are now separated in the events list
• Many UI fixes and optimizations

Breaking Changes

• Default values have been changed for motion detection. If you have specific values set in your config, it is recommended to remove them and re-calibrate as necessary.
• Stationary object validation is now required, anyone who manually set interval: 0 will need to delete this field from the config or increase the value
• Changes have been made to the DB schema, downgrading to 0.12 will require restoring from a backup or deleting the DB and existing recordings / snapshots.
  NOTE: Some users have reported migrations taking a while and frigate frontend is not available until the migrations are complete.
• New location for frigate.db: Due to the support for network shares in HA OS and the popularity of storing recordings on a NAS in general, the database has a new default location of /config/frigate.db. This change is going to be done for existing users too, so frigate will automatically move the db for you.

  • For Frigate Addon Users:
    The migration is handled automatically and no action is required from the user.

  • For Docker Container Users:
    The migration is handled for you, but you must make sure that the docker CLI or docker-compose binds the entire /config directory and not just the /config/config.yml file.

    For example, if your current volume mapping is - /host/path/config_folder/config.yml:/config/config.yml it should now be - /host/path/config_folder/:/config/

• The previously deprecated record -> retain_days config has been removed, this should be migrated to record -> retain -> days if you have not already.
• HomeAssistant Integration 5.0 beta is required for the new features introduced in this release.
• TensorRT has been updated and the model generation has been refactored. New models will need to be generated for 0.13 and have a new path, check the detector docs for more info.
• TensorRT now requires Nvidia Driver Version 530+
• The cameras in /stats have been moved to their own cameras block. This only affects users that were directly reading from the /stats data via the API or MQTT.

Frigate+

Frigate+ Model Launch

Frigate+ models are now rolling out by invitation. If you would like to be invited for early access to purchase a subscription, please send an email to blake@frigate.video from the email address on your Frigate+ account. For more information, see this discussion.

False positive submissions

You can now submit false positive detections to Frigate+ for training feedback. Note that you will still need to login to your Frigate+ account and verify the images to ensure all relevant objects are labeled before the image will be used during training. Confirming the object is a true positive will also upload the image.

Models

Frigate+ models are almost ready for launch, and this release has support for automatically downloading from Frigate+. The model and related information are stored at /config/model_cache. Once Frigate+ models are launched, you will be able to use the model id in your config.

model:
  path: plus://e63b7345cc83a84ed79dedfc99c16616

Community Supported Boards

In version 0.12 Frigate gained support for Intel iGPUs via OpenVINO and for Nvidia GPUs, this was a great step forward for compatibility and for more users being able to use Frigate effectively. There are many other devices, including dedicated SBCs (Small Board Computers), that Frigate could support but the maintainers do not have the time or hardware to support.

The community supported boards framework will allow community members who want to contribute to create and manage support for a particular board. This will allow Frigate to support more devices while allowing the maintainers to focus on improving the core of Frigate.

Nvidia Jetson Support

@madsciencetist is the first to contribute a community supported board build by adding support for Jetson 4.6 and Jetson 5 devices!

More information can be found in the detector docs on how to set this up.

https://deploy-preview-6262--frigate-docs.netlify.app/configuration/object_detectors#nvidia-tensorrt-detector

Object Tracking & Motion Detection

There have been many improvements to object tracking and motion detection which make Frigate more efficient and can reduce false positives.

New Object Tracker

Frigate now uses Norfair for object tracking, this means that object attributes like size, width/height ratio, and position are used to intelligently track objects. ID swapping between tracked objects is greatly reduced in situations where a car in the driveway was confused with a car passing in the street.

Improved Motion Detection

Frigate's motion detection is now more efficient. Motion detection will also re-calibrate for flashes like lightning or when the camera switches from color to IR mode or it is moved via PTZ. This reduces CPU usage during these changes and reduces false positives caused by excessive motion.

Historical Data

Frigate now uses the recordings timeline data to generate an 8x8 grid for each camera. This grid holds the range of expected regions (portions of the image sent to detection) for that area of the camera frame. This helps to improve accuracy of detection after motion has occurred and can reduce false positives in some cases.

The grid is updated on each startup and every day at 2 AM for each camera that is enabled.

Minimum Time For Zone Presence

Frigate now has a configurable field for the minimum amount of time an object needs to be in a zone before it is considered in a zone. For example, currently in Frigate if you have a front_yard zone and someone is walking by and briefly steps into the front yard Frigate would create an event for this. However, you may prefer for events to only be created if a person is in the front yard for multiple frames before creating an event.

zones:
  my_zone:
    inertia: 3  # <- 3 consecutive frames needed to consider object in zone
    coordinates: ...

Recordings Improvements

There have been a number of improvements to the management of recordings

Recordings Exporting

Frigate now supports exporting recordings in standard format or as a time-lapse. Exported recordings are stored in /media/frigate/exports and are downloadable from the UI.

Recordings Timeline Metadata

Frigate now saves metadata for key moments in an event such as when the tracked object is detected, enters or exits zone, becomes active / stationary,...