[Config Support]: Understanding Image Embedding stats to see what i'm doing wrong #22544

WorldOfGZ · 2026-03-20T08:09:47Z

WorldOfGZ
Mar 20, 2026

Describe the problem you are having

Hello

I'm not sure this is an issue, but interference time for Image Embedding seems really important (around 800ms) and seems not be linked to specific others things in metrics.

There is probably something wrong on my usage, but I can't find see it using documentation.
Could you please says me if this embedding time (which already goes to 1.2s) can cause issue on detection ? Specially false positive ?

Thanks

Version

0.17.0-f0d69f7

Frigate config file

mqtt:
  enabled: true
  host: 192.168.68.114
  user: '{FRIGATE_MQTT_USER}'
  password: '{FRIGATE_MQTT_PASS}'
  topic_prefix: frigate

tls:
  enabled: false

ffmpeg:
  hwaccel_args: preset-intel-qsv-h264   #Add after migration 14.1 -> 15.0 (before preset-vaapi)

record:
  enabled: true
  alerts:
    retain:
      days: 10
      mode: motion
    pre_capture: 2
    post_capture: 5
  detections:
    retain:
      days: 10
      mode: motion
    pre_capture: 2
    post_capture: 5
  continuous:
    days: 3
  motion:
    days: 3
snapshots:
  enabled: true
  clean_copy: true
  # Optional: Camera override for retention settings (default: global values)
  retain:
    # Required: Default retention days (default: shown below)
    default: 10

detectors:
  coral:
    type: edgetpu
    device: usb
detect:
  enabled: true
motion:
  enabled: true

#Frigate+ paid suscription
model:
  path: plus://xxx

#Go2RTC
go2rtc:
  streams:
    jardin_haut: #2560x1920
      - ffmpeg:http://192.168.68.113/flv?port=1935&app=bcs&stream=channel0_main.bcs&user={FRIGATE_CRED_REO_USER}&password={FRIGATE_CRED_REO_PASS}#video=copy#audio=copy#audio=opus
    jardin_haut_sub: #640x480
      - ffmpeg:http://192.168.68.113/flv?port=1935&app=bcs&stream=channel0_ext.bcs&user={FRIGATE_CRED_REO_USER}&password={FRIGATE_CRED_REO_PASS}
    jardin_bas: #2560x1920
      - ffmpeg:http://192.168.68.150/flv?port=1935&app=bcs&stream=channel0_main.bcs&user={FRIGATE_CRED_REO_USER}&password={FRIGATE_CRED_REO_PASS}#video=copy#audio=copy#audio=opus
    jardin_bas_sub: #640x480
      - ffmpeg:http://192.168.68.150/flv?port=1935&app=bcs&stream=channel0_ext.bcs&user={FRIGATE_CRED_REO_USER}&password={FRIGATE_CRED_REO_PASS}
    garage: #2560x1920
      - ffmpeg:http://192.168.68.101/flv?port=1935&app=bcs&stream=channel0_main.bcs&user={FRIGATE_CRED_REO_USER}&password={FRIGATE_CRED_REO_PASS}#video=copy#audio=copy#audio=opus
    garage_sub: #640x480
      - ffmpeg:http://192.168.68.101/flv?port=1935&app=bcs&stream=channel0_ext.bcs&user={FRIGATE_CRED_REO_USER}&password={FRIGATE_CRED_REO_PASS}
    grenier1: #2560x1920
      - ffmpeg:http://192.168.68.141/flv?port=1935&app=bcs&stream=channel0_main.bcs&user={FRIGATE_CRED_REO_USER}&password={FRIGATE_CRED_REO_PASS}#video=copy#audio=copy#audio=opus
    grenier1_sub: #640x480
      - ffmpeg:http://192.168.68.141/flv?port=1935&app=bcs&stream=channel0_ext.bcs&user={FRIGATE_CRED_REO_USER}&password={FRIGATE_CRED_REO_PASS}
    grenier2: #2560x1920
      - ffmpeg:http://192.168.68.143/flv?port=1935&app=bcs&stream=channel0_main.bcs&user={FRIGATE_CRED_REO_USER}&password={FRIGATE_CRED_REO_PASS}#video=copy#audio=copy#audio=opus
    grenier2_sub: #640x480
      - ffmpeg:http://192.168.68.143/flv?port=1935&app=bcs&stream=channel0_ext.bcs&user={FRIGATE_CRED_REO_USER}&password={FRIGATE_CRED_REO_PASS}
    grenier3: #2560x1920
      - ffmpeg:http://192.168.68.144/flv?port=1935&app=bcs&stream=channel0_main.bcs&user={FRIGATE_CRED_REO_USER}&password={FRIGATE_CRED_REO_PASS}#video=copy#audio=copy#audio=opus
    grenier3_sub: #640x480
      - ffmpeg:http://192.168.68.144/flv?port=1935&app=bcs&stream=channel0_ext.bcs&user={FRIGATE_CRED_REO_USER}&password={FRIGATE_CRED_REO_PASS}
    sonnette: #2560x1920
      - ffmpeg:http://192.168.68.151/flv?port=1935&app=bcs&stream=channel0_main.bcs&user={FRIGATE_CRED_REO_USER}&password={FRIGATE_CRED_REO_PASS}#video=copy#audio=copy#audio=opus
      - rtsp://{FRIGATE_CRED_REO_USER}:{FRIGATE_CRED_REO_PASS}@192.168.68.151:554/Preview_01_sub
    sonnette_sub: #640x480
      - ffmpeg:http://192.168.68.151/flv?port=1935&app=bcs&stream=channel0_ext.bcs&user={FRIGATE_CRED_REO_USER}&password={FRIGATE_CRED_REO_PASS}

  webrtc:
    candidates:
      - 192.168.68.119:8555
      - stun:8555
  ffmpeg:
    bin: /usr/lib/ffmpeg/7.0/bin/ffmpeg   #Add after migration 14.1 -> 15.0 (before ffmpeg)
    volume: -af "volume=30dB"

# Include all cameras by default in Birdseye view
birdseye:
  enabled: true
  mode: motion

#Global Object Settings
objects:
  track:
    - person
    - cat
    - dog
    - package #Frigate+ Suscription
    - face #Frigate+ Suscription
    - skunk #Frigate+ Suscription
    - rodent #Frigate+ Suscription
    - door #Frigate+ Suscription
  filters:
    person:
      min_area: 5000
      #max_area: 180000
      # Optional: minimum score for the object to initiate tracking (default: 0.6)
      min_score: 0.65
      # Optional: minimum decimal percentage for tracked object's computed score to be considered a true positive (default: 0.7)
      threshold: 0.85
    package: #Bigger due to fine tune
      min_score: .50 #Prev.65 - Score à partir duquel un objet sera suivi en détection
      threshold: .7 #Prev.9 - Score pour à partir duquel l'objet suivi est un vrai positif
  # Optional: Configuration for AI generated tracked object descriptions
  genai:
    # Optional: Enable AI object description generation (default: shown below)
    enabled: false
    # Optional: Use the object snapshot instead of thumbnails for description generation (default: shown below)
    use_snapshot: false
    # Optional: The default prompt for generating descriptions. Can use replacement
    # variables like "label", "sub_label", "camera" to make more dynamic. (default: shown below)
    prompt: Analyze the {label} in these images from the {camera} security 
      camera. Focus on the actions and potential intent of the {label}. If label
      is on_demand, just describe what you see on these images. No remarks on 
      the quality and length of the images or video. No remarks on potential 
      improvment. Make a complete comment with maximum 100 words, in french 
      language.
    # Optional: Object specific prompts to customize description results
    # Format: {label}: {prompt}
    object_prompts:
      person: You are tasked with analyzing footage captured by a security 
        camera. The camera is located outside the house. Examine the person in 
        the footage. What are they doing, and how might their actions suggest 
        their intent (e.g., delivering something, approaching, leaving, picking 
        something)? Only if they are carrying or interacting with an object, 
        indicate it with more details. Do not describe the background. No 
        remarks on the quality and length of the images or video. No remarks on 
        potential improvment. Make a complete comment with maximum 100 words, in
        french language.
      cat: You are tasked with analyzing footage captured by a security camera. 
        The camera is located outside the house. Observe the cat in these 
        images. Focus on its movement and intent (e.g., wandering, hunting, 
        interacting with objects of flowers). If the cat is near the bin or 
        engaging in any specific actions, mention it. Do not describe the 
        background. No remarks on the quality and length of the images or video.
        No remarks on potential improvment. If it's note a cat, mention it, and 
        explain what you see. Make a complete comment with maximum 100 words, in
        french language.
      dog: You are tasked with analyzing footage captured by a security camera. 
        The camera is located outside the house. Observe the dog in these 
        images. Focus on its movement and intent (e.g., wandering, hunting, 
        interacting with objects of flowers). If the dog is near the bin or 
        engaging in any specific actions, mention it. Do not describe the 
        background. No remarks on the quality and length of the images or video.
        No remarks on potential improvment. If it's note a dog, mention it, and 
        explain what you see.  Make a complete comment with maximum 100 words, 
        in french language.
    # Optional: objects to generate descriptions for (default: all objects that are tracked)
    objects:
      - person
      - cat
      - dog
    # Optional: Restrict generation to objects that entered any of the listed zones (default: none, all zones qualify)
    required_zones: []
    # Optional: What triggers to use to send frames for a tracked object to generative AI (default: shown below)
    send_triggers:
      # Once the object is no longer tracked
      tracked_object_end: true
      # Optional: After X many significant updates are received (default: shown below)
      after_significant_updates: 1
    # Optional: Save thumbnails sent to generative AI for review/debugging purposes (default: shown below)
    debug_save_thumbnails: false

# Optional: Review configuration
# NOTE: Can be overridden at the camera level
review:
  # Optional: alerts configuration
  alerts:
    # Optional: enables alerts for the camera (default: shown below)
    enabled: true
    # Optional: labels that qualify as an alert (default: shown below)
    labels:
      - person
    # Time to cutoff alerts after no alert-causing activity has occurred (default: shown below)
    cutoff_time: 40
    # Optional: required zones for an object to be marked as an alert (default: none)
    # NOTE: when settings required zones globally, this zone must exist on all cameras
    #       or the config will be considered invalid. In that case the required_zones
    #       should be configured at the camera level.
    #required_zones:
    #  - driveway
  # Optional: detections configuration
  detections:
    # Optional: enables detections for the camera (default: shown below)
    enabled: true
    # Optional: labels that qualify as a detection (default: all labels that are tracked / listened to)
    labels:
      - person
    # Time to cutoff detections after no detection-causing activity has occurred (default: shown below)
    cutoff_time: 30
    # Optional: required zones for an object to be marked as a detection (default: none)
    # NOTE: when settings required zones globally, this zone must exist on all cameras
    #       or the config will be considered invalid. In that case the required_zones
    #       should be configured at the camera level.
    #required_zones:
    #  - driveway
  # Optional: GenAI Review Summary Configuration
  genai:
    # Optional: Enable the GenAI review summary feature (default: shown below)
    enabled: false
    # Optional: Enable GenAI review summaries for alerts (default: shown below)
    alerts: false
    # Optional: Enable GenAI review summaries for detections (default: shown below)
    detections: false
    # Optional: Activity Context Prompt to give context to the GenAI what activity is and is not suspicious.
    # It is important to be direct and detailed. See documentation for the default prompt structure.
    activity_context_prompt: |
      ### Normal Activity Indicators (Level 0)
      - Known/verified people in any zone at any time
      - People with pets in residential areas
      - Deliveries or services during daytime/evening (6 AM - 10 PM): carrying packages to doors/porches, placing items, leaving
      - Services/maintenance workers with visible tools, uniforms, or service vehicles during daytime
      - Activity confined to public areas only (sidewalks, streets) without entering property at any time

      ### Suspicious Activity Indicators (Level 1)
      - **Testing or attempting to open doors/windows/handles on vehicles or buildings** — ALWAYS Level 1 regardless of time or duration
      - **Unidentified person in private areas (driveways, near vehicles/buildings) during late night/early morning (11 PM - 5 AM)** — ALWAYS Level 1 regardless of activity or duration
      - Taking items that don't belong to them (packages, objects from porches/driveways)
      - Climbing or jumping fences/barriers to access property
      - Attempting to conceal actions or items from view
      - Prolonged loitering: remaining in same area without visible purpose throughout most of the sequence

      ### Critical Threat Indicators (Level 2)
      - Holding break-in tools (crowbars, pry bars, bolt cutters)
      - Weapons visible (guns, knives, bats used aggressively)
      - Forced entry in progress
      - Physical aggression or violence
      - Active property damage or theft in progress

      ### Assessment Guidance
      Evaluate in this order:

      1. **If person is verified/known** → Level 0 regardless of time or activity
      2. **If person is unidentified:**
        - Check time: If late night/early morning (11 PM - 5 AM) AND in private areas (driveways, near vehicles/buildings) → Level 1
        - Check actions: If testing doors/handles, taking items, climbing → Level 1
        - Otherwise, if daytime/evening (6 AM - 10 PM) with clear legitimate purpose (delivery, service worker) → Level 0
      3. **Escalate to Level 2 if:** Weapons, break-in tools, forced entry in progress, violence, or active property damage visible (escalates from Level 0 or 1)

      The mere presence of an unidentified person in private areas during late night hours is inherently suspicious and warrants human review, regardless of what activity they appear to be doing or how brief the sequence is.
    # Optional: Image source for GenAI (default: preview)
    # Options: "preview" (uses cached preview frames at ~180p) or "recordings" (extracts frames from recordings at 480p)
    # Using "recordings" provides better image quality but uses more tokens per image.
    # Frame count is automatically calculated based on context window size, aspect ratio, and image source (capped at 20 frames).
    image_source: preview
    # Optional: Additional concerns that the GenAI should make note of (default: None)
    additional_concerns:
      - Animals in the garden
      - Package delivery
    # Optional: Preferred response language (default: English)
    preferred_language: French

face_recognition:
  # Optional: Enable face recognition
  enabled: true
  # Optional: Minimum face distance score required to mark as a potential match (default: 0.8)
  unknown_score: 0.8
  # Optional: Minimum face detection score required to detect a face (default: 0.7)
  # NOTE: This only applies when not running a Frigate+ model
  detection_threshold: 0.5
  # Optional: Minimum face distance score required to be considered a match (default: 0.9)
  recognition_threshold: 0.9
  # Optional: Min area of detected face box to consider running face recognition (default: 500)
  min_area: 500
  # Optional: Min face recognitions for the sub label to be applied to the person object (default: 1)
  min_faces: 2
  # Optional: Number of images of recognized faces to save for training (default: 100)
  save_attempts: 100
  # Optional: Apply a blur quality filter to adjust confidence based on the blur level of the image (default: True)
  blur_confidence_filter: true
  # Optional: Set the model size used face recognition.
  model_size: large

#License plate recognition
lpr:
  enabled: false

#Birds classification
classification:
  bird:
    enabled: false

  custom:
    Portillon garage:
      enabled: true
      name: Portillon garage
      threshold: 0.8
      state_config:
        cameras:
          garage:
            crop:
              - 0.45696042616762067
              - 0.14291288292821452
              - 0.7050733884865286
              - 0.47373016602009177
        motion: true
    Poubelles:
      enabled: true
      name: Poubelles
      threshold: 0.8
      state_config:
        cameras:
          sonnette:
            crop:
              - 0.44839383942404476
              - 0.618964882819448
              - 0.5063167338788129
              - 0.696195408759139
        motion: true
    Type colis:
      enabled: true
      name: Type colis
      threshold: 0.8
      object_config:
        objects:
          - package
        classification_type: sub_label
logger:
  default: info
#  logs:
#    frigate.video: debug  # Enable FFmpeg debug logs

##########################
#Camera specifications.
#The streams are defined in Go2Rtc
##########################
cameras:
  ##################################
  # JARDIN HAUT
  # REOLINK RLC-510A
  ##################################
  jardin_haut:
    ffmpeg:
      inputs:
        # High Resolution Stream
        - path: rtsp://127.0.0.1:8554/jardin_haut
          input_args: preset-rtsp-restream
          roles:
            - record
        # Low Resolution Stream
        - path: rtsp://127.0.0.1:8554/jardin_haut_sub
          input_args: preset-rtsp-restream
          roles:
            - detect
    live:
      streams:
        Main Stream: jardin_haut
        Sub Stream: jardin_haut_sub
    birdseye:
      mode: objects
    objects:
      track:
        - person
        - cat
        - dog
      filters:
        cat:
          max_area: 7500
    motion:
      mask:
        - 0,0,0.043,0,0,0.097
        - 0.74,0.055,0.741,0.012,0.272,0.012,0.272,0.058
      threshold: 50
      contour_area: 10
      improve_contrast: true
    zones:
      terrasse:
        coordinates: 
          0.175,0.578,0.169,0.525,0.041,0.57,0.079,0.689,0,0.719,0,1,1,1,1,0.825,1,0.673,0.957,0.763,0.838,0.582,0.719,0.418,0.717,0.352
        inertia: 3
        loitering_time: 0
      jardin:
        coordinates: 
          0.349,0.343,0.147,0.43,0.174,0.52,0.039,0.564,0,0.499,0,0,0.421,0,0.427,0.08,0.487,0.166,0.578,0.266
        inertia: 3
        loitering_time: 0
      toit:
        coordinates: 
          0.912,0.171,1,0.251,1,0.608,0.86,0.435,0.673,0.232,0.847,0.118
        inertia: 3
        loitering_time: 0
    detect:
      width: 640
      height: 480
      annotation_offset: 0 #Prev -1400
    ui:
      # Optional: Adjust sort order of cameras in the UI. Larger numbers come later
      order: 2
      # Optional: Whether or not to show the camera in the Frigate UI (default: shown below)
      dashboard: true
    notifications:
      enabled: false


  ##################################
  # JARDIN BAS
  # REOLINK RLC-510A
  ##################################
  jardin_bas:
    ffmpeg:
      inputs:
        #High Resolution Stream
        - path: rtsp://127.0.0.1:8554/jardin_bas
          input_args: preset-rtsp-restream
          roles:
            - record
        #Low Resolution Stream
        - path: rtsp://127.0.0.1:8554/jardin_bas_sub
          input_args: preset-rtsp-restream
          roles:
            - detect
    live:
      streams:
        Main Stream: jardin_bas
        Sub Stream: jardin_bas_sub
    objects:
      track:
        - person
        - cat
        - dog
      #filters:
      #  person:
      #    threshold: 0.7
      filters:
        cat:
          mask: 0.022,0,0.022,0.059,0.068,0.054,0.067,0
    birdseye:
      mode: objects
    motion:
      mask:
        - 0,0.285,0.076,0.596,0.159,0.837,0.199,0.82,0.139,0.537,0.019,0,0,0
        - 0,1,0,0.432,0.07,0.77,0.104,1
        - 0.731,0.016,0.73,0.059,0.283,0.058,0.284,0.009
        - 1,0.058,0.775,0.002,1,0.001
      threshold: 53
      contour_area: 15
      improve_contrast: true
    detect:
      #Force another resolution so that the GPU detects the faces
      width: 640
      height: 480
      annotation_offset: 0 # Previous-1600
    zones:
      terrasse:
        coordinates: 
          0.076,0.309,0.118,0.64,0.207,1,1,1,1,0.381,0.498,0.166,0.303,0.225
        inertia: 3
        loitering_time: 0
    ui:
      # Optional: Adjust sort order of cameras in the UI. Larger numbers come later
      order: 1
      # Optional: Whether or not to show the camera in the Frigate UI (default: shown below)
      dashboard: true
    notifications:
      enabled: false


  ##################################
  # GARAGE
  # REOLINK RLC-510A
  ##################################
  garage:
    ffmpeg:
      inputs:
        #High Resolution Stream
        #Use main stream due to face detection
        - path: rtsp://127.0.0.1:8554/garage
          input_args: preset-rtsp-restream
          roles:
            - record
            - detect
    live:
      streams:
        Main Stream: garage
        Sub Stream: garage_sub
    objects:
      track:
        - person
        - face #Frigate+ Suscription
        - door #Frigate+ Suscription
    birdseye:
      mode: objects
    detect:
      #Force another resolution so that the GPU detects the faces
      width: 1280
      height: 960
      annotation_offset: -500 #Prev -1000
    motion:
      threshold: 60
      contour_area: 20
      improve_contrast: true
      mask:
        - 0.344,0.041,0.658,0.04,0.657,0.002,0.344,0.001
        - 0.924,0.184,0.905,0.529,0.956,0.756,1,0.536,1,0.273
    ui:
      # Optional: Adjust sort order of cameras in the UI. Larger numbers come later
      order: 3
      # Optional: Whether or not to show the camera in the Frigate UI (default: shown below)
      dashboard: true
    notifications:
      enabled: false


  ##################################
  # GRENIER 1
  # Reolink RLC-520A
  ##################################
  grenier1:
    ffmpeg:
      output_args:
        record: preset-record-generic-audio-aac
      inputs:
        #High Resolution Stream
        #Use main stream due to rodent detection
        - path: rtsp://127.0.0.1:8554/grenier1
          input_args: preset-rtsp-restream
          roles:
            - record
            - detect
    live:
      streams:
        Main Stream: grenier1
        Sub Stream: grenier1_sub
    objects:
      track:
        - cat
        - dog
        - skunk #Frigate+ Suscription
        - rodent #Frigate+ Suscription
      genai:
        enabled: false
    birdseye:
      mode: objects
    motion:
      mask: 0.341,0.042,0.661,0.042,0.664,0,0.338,0
      threshold: 45
      contour_area: 15
      improve_contrast: true
    detect:
      #Force another resolution so that the GPU detects the faces
      width: 1280
      height: 960
    ui:
      # Optional: Adjust sort order of cameras in the UI. Larger numbers come later
      order: 4
      # Optional: Whether or not to show the camera in the Frigate UI (default: shown below)
      dashboard: true
    face_recognition:
      enabled: false
    notifications:
      enabled: false

  ##################################
  # GRENIER 2
  # Reolink RLC-520A
  ##################################
  grenier2:
    ffmpeg:
      output_args:
        record: preset-record-generic-audio-aac
      inputs:
        #High Resolution Stream
        #Use main stream due to rodent detection
        - path: rtsp://127.0.0.1:8554/grenier2
          input_args: preset-rtsp-restream
          roles:
            - record
            - detect
    live:
      streams:
        Main Stream: grenier2
        Sub Stream: grenier2_sub
    objects:
      track:
        - cat
        - dog
        - skunk #Frigate+ Suscription
        - rodent #Frigate+ Suscription
      filters:
        skunk: {}
      genai:
        enabled: false
    birdseye:
      mode: objects
    motion:
      mask:
        - 0.336,0.04,0.673,0.04,0.673,0,0.334,0
        - 0.168,0.303,0,0.303,0,1,0.114,1
      threshold: 40
      contour_area: 15
      improve_contrast: true
    detect:
      width: 1280
      height: 960
    ui:
      # Optional: Adjust sort order of cameras in the UI. Larger numbers come later
      order: 5
      # Optional: Whether or not to show the camera in the Frigate UI (default: shown below)
      dashboard: true
    face_recognition:
      enabled: false
    notifications:
      enabled: false


  ##################################
  # GRENIER 3
  # Reolink RLC-520A
  ##################################
  grenier3:
    ffmpeg:
      output_args:
        record: preset-record-generic-audio-aac
      inputs:
        #High Resolution Stream
        #Use main stream due to rodent detection
        - path: rtsp://127.0.0.1:8554/grenier3
          input_args: preset-rtsp-restream
          roles:
            - record
            - detect
    live:
      streams:
        Main Stream: grenier3
        Sub Stream: grenier3_sub
    objects:
      track:
        - cat
        - dog
        - skunk #Frigate+ Suscription
        - rodent #Frigate+ Suscription
      filters:
        skunk: {}
      genai:
        enabled: false
    birdseye:
      mode: objects
    motion:
      mask: 0.33,0.039,0.668,0.039,0.67,0,0.33,0
      threshold: 40
      contour_area: 10
      improve_contrast: true
    detect:
      width: 1280
      height: 960
    ui:
      # Optional: Adjust sort order of cameras in the UI. Larger numbers come later
      order: 6
      # Optional: Whether or not to show the camera in the Frigate UI (default: shown below)
      dashboard: true
    face_recognition:
      enabled: false
    notifications:
      enabled: false





  ##################################
  # SONNETTE
  # Reolink Video Doorbell PoE
  ##################################
  sonnette:
    ffmpeg:
      inputs:
        #High resolution stream
        #Use main stream due to face detection
        - path: rtsp://127.0.0.1:8554/sonnette
          input_args: preset-rtsp-restream
          roles:
            - record
            - detect
      output_args:
        record: preset-record-generic-audio-aac #Add after migration 16.0 (before preset-record-generic-audio-copy)
    live:
      streams:
        Main Stream: sonnette
        Sub Stream: sonnette_sub
    birdseye:
      mode: objects
    objects:
      track:
        - person
        - cat
        - dog
        - package #Frigate+ Suscription
        - face #Frigate+ Suscription
      filters:
        person:
          min_area: 7500
          threshold: 0.7
        #cat:
        #  max_area: 20000
        #dog:
        #  max_area: 22000
    motion:
      mask:
        - 0.312,0,0.682,0,0.684,0.047,0.312,0.042
        - 0.434,0.281,0.66,0.299,0.655,0.622,0.444,0.629
      threshold: 60
      contour_area: 20
      improve_contrast: true
    detect:
      #Force another resolution so that the GPU detects the faces
      width: 1280
      height: 960
      annotation_offset: -2500 #(Value when stream sub used : -2550)
    zones:
      jardin:
        coordinates: 0.444,0.633,0.656,0.627,0.652,0.749,0.644,0.818,0.448,0.838
        loitering_time: 0
        inertia: 3
      entree:
        coordinates: 
          0.445,0.843,0.43,0.225,0.001,0.192,0.001,0.997,0.999,0.997,0.997,0.253,0.666,0.259,0.656,0.743,0.646,0.822
        loitering_time: 0
        inertia: 3
    review:
      alerts:
        required_zones:
          - entree
          - jardin
    # Optional: Configuration for how camera is handled in the GUI.
    ui:
      # Optional: Adjust sort order of cameras in the UI. Larger numbers come later
      order: 0
      # Optional: Whether or not to show the camera in the Frigate UI (default: shown below)
      dashboard: true
    notifications:
      enabled: false



version: 0.17-0

camera_groups:
  Grenier:
    order: 2
    icon: LuSquirrel
    cameras:
      - grenier1
      - grenier2
      - grenier3
  Extérieurs:
    order: 1
    icon: LuTrees
    cameras:
      - garage
      - jardin_bas
      - jardin_haut
      - sonnette

semantic_search:
  enabled: true
  reindex: false
  model: "jinav1"
  model_size: small

genai:
  provider: gemini
  api_key: '{FRIGATE_GEMINI_API_KEY}'
  model: gemini-2.5-flash

notifications:
  enabled: false
  email: xxx

telemetry:
  stats:
    intel_gpu_device: drm:/dev/dri/card0

Relevant Frigate log output

no errors

Relevant go2rtc log output

no errors

Frigate stats

{
  "cameras": {
    "jardin_haut": {
      "camera_fps": 5.1,
      "process_fps": 5.1,
      "skipped_fps": 0,
      "detection_fps": 0,
      "detection_enabled": false,
      "pid": 963,
      "capture_pid": 1022,
      "ffmpeg_pid": 1072,
      "audio_rms": 0,
      "audio_dBFS": 0
    },
    "jardin_bas": {
      "camera_fps": 5.1,
      "process_fps": 5.1,
      "skipped_fps": 0,
      "detection_fps": 0,
      "detection_enabled": false,
      "pid": 1136,
      "capture_pid": 1197,
      "ffmpeg_pid": 1244,
      "audio_rms": 0,
      "audio_dBFS": 0
    },
    "garage": {
      "camera_fps": 5.1,
      "process_fps": 5.1,
      "skipped_fps": 0,
      "detection_fps": 0,
      "detection_enabled": true,
      "pid": 1268,
      "capture_pid": 1345,
      "ffmpeg_pid": 1398,
      "audio_rms": 0,
      "audio_dBFS": 0
    },
    "grenier1": {
      "camera_fps": 5.1,
      "process_fps": 5.1,
      "skipped_fps": 0,
      "detection_fps": 0,
      "detection_enabled": true,
      "pid": 1409,
      "capture_pid": 1487,
      "ffmpeg_pid": 1540,
      "audio_rms": 0,
      "audio_dBFS": 0
    },
    "grenier2": {
      "camera_fps": 5,
      "process_fps": 5,
      "skipped_fps": 0,
      "detection_fps": 0,
      "detection_enabled": true,
      "pid": 1551,
      "capture_pid": 1630,
      "ffmpeg_pid": 1680,
      "audio_rms": 0,
      "audio_dBFS": 0
    },
    "grenier3": {
      "camera_fps": 5,
      "process_fps": 5,
      "skipped_fps": 0,
      "detection_fps": 0,
      "detection_enabled": true,
      "pid": 1694,
      "capture_pid": 1773,
      "ffmpeg_pid": 1825,
      "audio_rms": 0,
      "audio_dBFS": 0
    },
    "sonnette": {
      "camera_fps": 5.1,
      "process_fps": 5.1,
      "skipped_fps": 0,
      "detection_fps": 0,
      "detection_enabled": true,
      "pid": 1837,
      "capture_pid": 1917,
      "ffmpeg_pid": 1963,
      "audio_rms": 0,
      "audio_dBFS": 0
    }
  },
  "detectors": {
    "coral": {
      "inference_speed": 5.99,
      "detection_start": 0,
      "pid": 870
    }
  },
  "camera_fps": 35.5,
  "process_fps": 35.5,
  "skipped_fps": 0,
  "detection_fps": 0,
  "embeddings": {
    "image_embedding_speed": 794.8,
    "image_embedding": 0.1,
    "text_embedding_speed": 0,
    "text_embedding": 0,
    "face_recognition_speed": 90.39,
    "face_recognition": 0,
    "Portillon garage_classification_speed": 3.92,
    "Portillon garage_classification_events_per_second": 0,
    "Poubelles_classification_speed": 3.78,
    "Poubelles_classification_events_per_second": 0,
    "Type colis_classification_speed": 0,
    "Type colis_classification_events_per_second": 0
  },
  "gpu_usages": {
    "intel-qsv": {
      "gpu": "0.0%",
      "mem": "-%",
      "clients": {
        "1825": "0.1%",
        "1244": "0.01%",
        "825": "0.0%",
        "1072": "0.0%",
        "1398": "0.0%",
        "1540": "0.0%",
        "1680": "0.0%",
        "1963": "0.0%"
      }
    }
  },
  "cpu_usages": {
    "frigate.full_system": {
      "cpu": "14.2",
      "mem": "68.2"
    },
    "1": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "/package/admin/s6/command/s6-svscan -d4 -- /run/service"
    },
    "17": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "s6-supervise s6-linux-init-shutdownd"
    },
    "18": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "/package/admin/s6-linux-init/command/s6-linux-init-shutdownd -d3 -c /run/s6/basedir -g 3000 -C -B"
    },
    "27": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "s6-supervise nginx-log"
    },
    "28": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "s6-supervise certsync"
    },
    "29": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "s6-supervise go2rtc"
    },
    "30": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "s6-supervise s6rc-oneshot-runner"
    },
    "31": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "s6-supervise frigate-log"
    },
    "32": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "s6-supervise go2rtc-healthcheck"
    },
    "33": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "s6-supervise nginx"
    },
    "34": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "s6-supervise certsync-log"
    },
    "35": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "s6-supervise s6rc-fdholder"
    },
    "36": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "s6-supervise frigate"
    },
    "37": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "s6-supervise go2rtc-log"
    },
    "48": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "/package/admin/s6/command/s6-ipcserverd -1 -- /package/admin/s6/command/s6-ipcserver-access -v0 -E -l0 -i data/rules -- /package/admin/s6/command/s6-sudod -t 30000 -- /package/admin/s6-rc/command/s6-rc-oneshot-run -l ../.. --"
    },
    "49": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "/package/admin/s6-2.13.2.0/command/s6-fdholderd -1 -n 256 -i data/rules"
    },
    "97": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "s6-log -b -- T 1 n0 s10000000 T /dev/shm/logs/nginx"
    },
    "98": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "s6-log -b -- T 1 n0 s10000000 T /dev/shm/logs/go2rtc"
    },
    "99": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "s6-log -b -- T 1 n0 s10000000 T /dev/shm/logs/certsync"
    },
    "100": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "s6-log -b -- T 1 n0 s10000000 T /dev/shm/logs/frigate"
    },
    "128": {
      "cpu": "3.5",
      "cpu_average": "3",
      "mem": "0.1",
      "cmdline": "/usr/local/go2rtc/bin/go2rtc -config=/config/go2rtc_homekit.yml -config=/dev/shm/go2rtc.yaml"
    },
    "135": {
      "cpu": "3.2",
      "cpu_average": "3",
      "mem": "1.5",
      "cmdline": "python3 -u -m frigate"
    },
    "140": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "bash ./run.user go2rtc-healthcheck"
    },
    "143": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "nginx: master process nginx"
    },
    "187": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "nginx: worker process"
    },
    "188": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "nginx: worker process"
    },
    "189": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "nginx: worker process"
    },
    "190": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "nginx: worker process"
    },
    "191": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "nginx: cache manager process"
    },
    "349": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "bash ./run.user certsync"
    },
    "370": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "sleep 9999"
    },
    "374": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "/usr/bin/python3 -Wignore:::numpy.core.getlimits -c from multiprocessing.resource_tracker import main;main(5)"
    },
    "375": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.7",
      "cmdline": "/usr/bin/python3 -Wignore:::numpy.core.getlimits -c from multiprocessing.forkserver import main; main(5, 7, ['sqlite3', 'numpy', 'cv2', 'peewee', 'zmq', 'ruamel.yaml', 'frigate.camera.maintainer'], **{'sys_path': ['/opt/frigate', '/usr/lib/python311.zip', '/usr/lib/python3.11', '/usr/lib/python3.11/lib-dynload', '/usr/local/lib/python3.11/dist-packages', '/usr/lib/python3/dist-packages']})"
    },
    "382": {
      "cpu": "8.2",
      "cpu_average": "8",
      "mem": "0.4",
      "cmdline": "/usr/bin/python3 -Wignore:::numpy.core.getlimits -c from multiprocessing.forkserver import main; main(5, 7, ['sqlite3', 'numpy', 'cv2', 'peewee', 'zmq', 'ruamel.yaml', 'frigate.camera.maintainer'], **{'sys_path': ['/opt/frigate', '/usr/lib/python311.zip', '/usr/lib/python3.11', '/usr/lib/python3.11/lib-dynload', '/usr/local/lib/python3.11/dist-packages', '/usr/lib/python3/dist-packages']})"
    },
    "802": {
      "cpu": "3.1",
      "cpu_average": "2",
      "mem": "0.5",
      "cmdline": "frigate.recording_manager                                                                                                                                                                                                                                                                                                                                                                               "
    },
    "815": {
      "cpu": "0.5",
      "cpu_average": "0",
      "mem": "0.5",
      "cmdline": "frigate.review_segment_manager                                                                                                                                                                                                                                                                                                                                                                          "
    },
    "825": {
      "cpu": "5.1",
      "cpu_average": "8",
      "mem": "5.2",
      "cmdline": "frigate.embeddings_manager                                                                                                                                                                                                                                                                                                                                                                              "
    },
    "870": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.5",
      "cmdline": "frigate.detector:coral                                                                                                                                                                                                                                                                                                                                                                                  "
    },
    "884": {
      "cpu": "0.9",
      "cpu_average": "0",
      "mem": "0.5",
      "cmdline": "frigate.output                                                                                                                                                                                                                                                                                                                                                                                          "
    },
    "911": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -threads 1 -f rawvideo -pix_fmt yuv420p -video_size 640x480 -i pipe: -threads 1 -f mpegts -s 960x720 -codec:v mpeg1video -q 8 -bf 0 pipe:"
    },
    "916": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -threads 1 -f rawvideo -pix_fmt yuv420p -video_size 640x480 -i pipe: -threads 1 -f mpegts -s 960x720 -codec:v mpeg1video -q 8 -bf 0 pipe:"
    },
    "921": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -threads 1 -f rawvideo -pix_fmt yuv420p -video_size 1280x960 -i pipe: -threads 1 -f mpegts -s 960x720 -codec:v mpeg1video -q 8 -bf 0 pipe:"
    },
    "926": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -threads 1 -f rawvideo -pix_fmt yuv420p -video_size 1280x960 -i pipe: -threads 1 -f mpegts -s 960x720 -codec:v mpeg1video -q 8 -bf 0 pipe:"
    },
    "931": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -threads 1 -f rawvideo -pix_fmt yuv420p -video_size 1280x960 -i pipe: -threads 1 -f mpegts -s 960x720 -codec:v mpeg1video -q 8 -bf 0 pipe:"
    },
    "936": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -threads 1 -f rawvideo -pix_fmt yuv420p -video_size 1280x960 -i pipe: -threads 1 -f mpegts -s 960x720 -codec:v mpeg1video -q 8 -bf 0 pipe:"
    },
    "943": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -threads 1 -f rawvideo -pix_fmt yuv420p -video_size 1280x960 -i pipe: -threads 1 -f mpegts -s 960x720 -codec:v mpeg1video -q 8 -bf 0 pipe:"
    },
    "951": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -threads 1 -f rawvideo -pix_fmt yuv420p -video_size 1280x720 -i pipe: -threads 1 -f mpegts -s 1280x720 -codec:v mpeg1video -q 8 -bf 0 pipe:"
    },
    "963": {
      "cpu": "0.1",
      "cpu_average": "0",
      "mem": "0.4",
      "cmdline": "frigate.process:jardin_haut                                                                                                                                                                                                                                                                                                                                                                             "
    },
    "1022": {
      "cpu": "0.4",
      "cpu_average": "0",
      "mem": "0.4",
      "cmdline": "frigate.capture:jardin_haut                                                                                                                                                                                                                                                                                                                                                                             "
    },
    "1072": {
      "cpu": "1.5",
      "cpu_average": "1",
      "mem": "0.2",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -loglevel warning -threads 2 -hwaccel qsv -qsv_device /dev/dri/renderD128 -hwaccel_output_format qsv -c:v h264_qsv -bsf:v dump_extra -user_agent FFmpeg Frigate/0.17.0-f0d69f7 -rtsp_transport tcp -timeout 10000000 -i rtsp://127.0.0.1:8554/jardin_haut_sub -r 5 -vf vpp_qsv=framerate=5:w=640:h=480:format=nv12,hwdownload,format=nv12,format=yuv420p -threads 2 -f rawvideo -pix_fmt yuv420p pipe:"
    },
    "1081": {
      "cpu": "1.9",
      "cpu_average": "1",
      "mem": "0.1",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -loglevel warning -threads 2 -user_agent FFmpeg Frigate/0.17.0-f0d69f7 -rtsp_transport tcp -timeout 10000000 -i rtsp://127.0.0.1:8554/jardin_haut -f segment -segment_time 10 -segment_format mp4 -reset_timestamps 1 -strftime 1 -c:v copy -c:a aac /tmp/cache/jardin_haut@%Y%m%d%H%M%S%z.mp4"
    },
    "1084": {
      "cpu": "2.0",
      "cpu_average": "1",
      "mem": "0.1",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -v error -fflags nobuffer -flags low_delay -i http://192.168.68.113/flv?port=1935&app=bcs&stream=channel0_main.bcs&user=*&password=* -map 0:v:0? -c:v:0 copy -map 0:a:0? -c:a:0 copy -map 0:a:0? -c:a:1 libopus -application:a:1 lowdelay -min_comp 0 -user_agent ffmpeg/go2rtc -rtsp_transport tcp -f rtsp rtsp://127.0.0.1:8554/548bf90be36b4b152e4e0d7f1a8eceb7"
    },
    "1120": {
      "cpu": "0.6",
      "cpu_average": "0",
      "mem": "0.1",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -v error -fflags nobuffer -flags low_delay -i http://192.168.68.113/flv?port=1935&app=bcs&stream=channel0_ext.bcs&user=*&password=* -c copy -user_agent ffmpeg/go2rtc -rtsp_transport tcp -f rtsp rtsp://127.0.0.1:8554/67a74fdbfbb6feaba72d01fcf8f6806b"
    },
    "1136": {
      "cpu": "0.1",
      "cpu_average": "0",
      "mem": "0.5",
      "cmdline": "frigate.process:jardin_bas                                                                                                                                                                                                                                                                                                                                                                              "
    },
    "1197": {
      "cpu": "0.4",
      "cpu_average": "0",
      "mem": "0.4",
      "cmdline": "frigate.capture:jardin_bas                                                                                                                                                                                                                                                                                                                                                                              "
    },
    "1244": {
      "cpu": "1.6",
      "cpu_average": "1",
      "mem": "0.2",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -loglevel warning -threads 2 -hwaccel qsv -qsv_device /dev/dri/renderD128 -hwaccel_output_format qsv -c:v h264_qsv -bsf:v dump_extra -user_agent FFmpeg Frigate/0.17.0-f0d69f7 -rtsp_transport tcp -timeout 10000000 -i rtsp://127.0.0.1:8554/jardin_bas_sub -r 5 -vf vpp_qsv=framerate=5:w=640:h=480:format=nv12,hwdownload,format=nv12,format=yuv420p -threads 2 -f rawvideo -pix_fmt yuv420p pipe:"
    },
    "1251": {
      "cpu": "2.2",
      "cpu_average": "1",
      "mem": "0.1",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -loglevel warning -threads 2 -user_agent FFmpeg Frigate/0.17.0-f0d69f7 -rtsp_transport tcp -timeout 10000000 -i rtsp://127.0.0.1:8554/jardin_bas -f segment -segment_time 10 -segment_format mp4 -reset_timestamps 1 -strftime 1 -c:v copy -c:a aac /tmp/cache/jardin_bas@%Y%m%d%H%M%S%z.mp4"
    },
    "1256": {
      "cpu": "2.5",
      "cpu_average": "2",
      "mem": "0.1",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -v error -fflags nobuffer -flags low_delay -i http://192.168.68.150/flv?port=1935&app=bcs&stream=channel0_main.bcs&user=*&password=* -map 0:v:0? -c:v:0 copy -map 0:a:0? -c:a:0 copy -map 0:a:0? -c:a:1 libopus -application:a:1 lowdelay -min_comp 0 -user_agent ffmpeg/go2rtc -rtsp_transport tcp -f rtsp rtsp://127.0.0.1:8554/4f8eeb80a88646b85dd4ca84546676a9"
    },
    "1268": {
      "cpu": "0.8",
      "cpu_average": "0",
      "mem": "0.4",
      "cmdline": "frigate.process:garage                                                                                                                                                                                                                                                                                                                                                                                  "
    },
    "1294": {
      "cpu": "0.6",
      "cpu_average": "0",
      "mem": "0.1",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -v error -fflags nobuffer -flags low_delay -i http://192.168.68.150/flv?port=1935&app=bcs&stream=channel0_ext.bcs&user=*&password=* -c copy -user_agent ffmpeg/go2rtc -rtsp_transport tcp -f rtsp rtsp://127.0.0.1:8554/6d86d07b01c72bc0212fd6a1cd775d66"
    },
    "1345": {
      "cpu": "0.8",
      "cpu_average": "0",
      "mem": "0.4",
      "cmdline": "frigate.capture:garage                                                                                                                                                                                                                                                                                                                                                                                  "
    },
    "1398": {
      "cpu": "3.7",
      "cpu_average": "3",
      "mem": "0.3",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -loglevel warning -threads 2 -hwaccel qsv -qsv_device /dev/dri/renderD128 -hwaccel_output_format qsv -c:v h264_qsv -bsf:v dump_extra -user_agent FFmpeg Frigate/0.17.0-f0d69f7 -rtsp_transport tcp -timeout 10000000 -i rtsp://127.0.0.1:8554/garage -f segment -segment_time 10 -segment_format mp4 -reset_timestamps 1 -strftime 1 -c:v copy -c:a aac /tmp/cache/garage@%Y%m%d%H%M%S%z.mp4 -r 5 -vf vpp_qsv=framerate=5:w=1280:h=960:format=nv12,hwdownload,format=nv12,format=yuv420p -threads 2 -f rawvideo -pix_fmt yuv420p pipe:"
    },
    "1409": {
      "cpu": "0.8",
      "cpu_average": "0",
      "mem": "0.5",
      "cmdline": "frigate.process:grenier1                                                                                                                                                                                                                                                                                                                                                                                "
    },
    "1486": {
      "cpu": "1.6",
      "cpu_average": "1",
      "mem": "0.1",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -v error -fflags nobuffer -flags low_delay -i http://192.168.68.101/flv?port=1935&app=bcs&stream=channel0_main.bcs&user=*&password=* -map 0:v:0? -c:v:0 copy -map 0:a:0? -c:a:0 copy -map 0:a:0? -c:a:1 libopus -application:a:1 lowdelay -min_comp 0 -user_agent ffmpeg/go2rtc -rtsp_transport tcp -f rtsp rtsp://127.0.0.1:8554/2e81b40bc84a7424e2d15b281df001b5"
    },
    "1487": {
      "cpu": "0.8",
      "cpu_average": "0",
      "mem": "0.4",
      "cmdline": "frigate.capture:grenier1                                                                                                                                                                                                                                                                                                                                                                                "
    },
    "1540": {
      "cpu": "4.9",
      "cpu_average": "4",
      "mem": "0.3",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -loglevel warning -threads 2 -hwaccel qsv -qsv_device /dev/dri/renderD128 -hwaccel_output_format qsv -c:v h264_qsv -bsf:v dump_extra -user_agent FFmpeg Frigate/0.17.0-f0d69f7 -rtsp_transport tcp -timeout 10000000 -i rtsp://127.0.0.1:8554/grenier1 -f segment -segment_time 10 -segment_format mp4 -reset_timestamps 1 -strftime 1 -c:v copy -c:a aac /tmp/cache/grenier1@%Y%m%d%H%M%S%z.mp4 -r 5 -vf vpp_qsv=framerate=5:w=1280:h=960:format=nv12,hwdownload,format=nv12,format=yuv420p -threads 2 -f rawvideo -pix_fmt yuv420p pipe:"
    },
    "1551": {
      "cpu": "0.7",
      "cpu_average": "0",
      "mem": "0.5",
      "cmdline": "frigate.process:grenier2                                                                                                                                                                                                                                                                                                                                                                                "
    },
    "1600": {
      "cpu": "2.0",
      "cpu_average": "1",
      "mem": "0.1",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -v error -fflags nobuffer -flags low_delay -i http://192.168.68.141/flv?port=1935&app=bcs&stream=channel0_main.bcs&user=*&password=* -map 0:v:0? -c:v:0 copy -map 0:a:0? -c:a:0 copy -map 0:a:0? -c:a:1 libopus -application:a:1 lowdelay -min_comp 0 -user_agent ffmpeg/go2rtc -rtsp_transport tcp -f rtsp rtsp://127.0.0.1:8554/3df47579e1151d15a80a53eabd22d934"
    },
    "1630": {
      "cpu": "0.9",
      "cpu_average": "0",
      "mem": "0.4",
      "cmdline": "frigate.capture:grenier2                                                                                                                                                                                                                                                                                                                                                                                "
    },
    "1680": {
      "cpu": "5.1",
      "cpu_average": "4",
      "mem": "0.3",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -loglevel warning -threads 2 -hwaccel qsv -qsv_device /dev/dri/renderD128 -hwaccel_output_format qsv -c:v h264_qsv -bsf:v dump_extra -user_agent FFmpeg Frigate/0.17.0-f0d69f7 -rtsp_transport tcp -timeout 10000000 -i rtsp://127.0.0.1:8554/grenier2 -f segment -segment_time 10 -segment_format mp4 -reset_timestamps 1 -strftime 1 -c:v copy -c:a aac /tmp/cache/grenier2@%Y%m%d%H%M%S%z.mp4 -r 5 -vf vpp_qsv=framerate=5:w=1280:h=960:format=nv12,hwdownload,format=nv12,format=yuv420p -threads 2 -f rawvideo -pix_fmt yuv420p pipe:"
    },
    "1694": {
      "cpu": "0.8",
      "cpu_average": "0",
      "mem": "0.5",
      "cmdline": "frigate.process:grenier3                                                                                                                                                                                                                                                                                                                                                                                "
    },
    "1768": {
      "cpu": "1.9",
      "cpu_average": "1",
      "mem": "0.1",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -v error -fflags nobuffer -flags low_delay -i http://192.168.68.143/flv?port=1935&app=bcs&stream=channel0_main.bcs&user=*&password=* -map 0:v:0? -c:v:0 copy -map 0:a:0? -c:a:0 copy -map 0:a:0? -c:a:1 libopus -application:a:1 lowdelay -min_comp 0 -user_agent ffmpeg/go2rtc -rtsp_transport tcp -f rtsp rtsp://127.0.0.1:8554/946287868c45644ae1b41f5d7d6649cd"
    },
    "1773": {
      "cpu": "0.8",
      "cpu_average": "0",
      "mem": "0.4",
      "cmdline": "frigate.capture:grenier3                                                                                                                                                                                                                                                                                                                                                                                "
    },
    "1825": {
      "cpu": "5.0",
      "cpu_average": "4",
      "mem": "0.3",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -loglevel warning -threads 2 -hwaccel qsv -qsv_device /dev/dri/renderD128 -hwaccel_output_format qsv -c:v h264_qsv -bsf:v dump_extra -user_agent FFmpeg Frigate/0.17.0-f0d69f7 -rtsp_transport tcp -timeout 10000000 -i rtsp://127.0.0.1:8554/grenier3 -f segment -segment_time 10 -segment_format mp4 -reset_timestamps 1 -strftime 1 -c:v copy -c:a aac /tmp/cache/grenier3@%Y%m%d%H%M%S%z.mp4 -r 5 -vf vpp_qsv=framerate=5:w=1280:h=960:format=nv12,hwdownload,format=nv12,format=yuv420p -threads 2 -f rawvideo -pix_fmt yuv420p pipe:"
    },
    "1837": {
      "cpu": "0.7",
      "cpu_average": "0",
      "mem": "0.5",
      "cmdline": "frigate.process:sonnette                                                                                                                                                                                                                                                                                                                                                                                "
    },
    "1894": {
      "cpu": "1.9",
      "cpu_average": "1",
      "mem": "0.1",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -v error -fflags nobuffer -flags low_delay -i http://192.168.68.144/flv?port=1935&app=bcs&stream=channel0_main.bcs&user=*&password=* -map 0:v:0? -c:v:0 copy -map 0:a:0? -c:a:0 copy -map 0:a:0? -c:a:1 libopus -application:a:1 lowdelay -min_comp 0 -user_agent ffmpeg/go2rtc -rtsp_transport tcp -f rtsp rtsp://127.0.0.1:8554/474f4e513c8d434854356694ca6c6598"
    },
    "1917": {
      "cpu": "0.8",
      "cpu_average": "0",
      "mem": "0.4",
      "cmdline": "frigate.capture:sonnette                                                                                                                                                                                                                                                                                                                                                                                "
    },
    "1963": {
      "cpu": "4.2",
      "cpu_average": "4",
      "mem": "0.3",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -loglevel warning -threads 2 -hwaccel qsv -qsv_device /dev/dri/renderD128 -hwaccel_output_format qsv -c:v h264_qsv -bsf:v dump_extra -user_agent FFmpeg Frigate/0.17.0-f0d69f7 -rtsp_transport tcp -timeout 10000000 -i rtsp://127.0.0.1:8554/sonnette -f segment -segment_time 10 -segment_format mp4 -reset_timestamps 1 -strftime 1 -c:v copy -c:a aac /tmp/cache/sonnette@%Y%m%d%H%M%S%z.mp4 -r 5 -vf vpp_qsv=framerate=5:w=1280:h=960:format=nv12,hwdownload,format=nv12,format=yuv420p -threads 2 -f rawvideo -pix_fmt yuv420p pipe:"
    },
    "1989": {
      "cpu": "1.5",
      "cpu_average": "1",
      "mem": "0.1",
      "cmdline": "/usr/lib/ffmpeg/7.0/bin/ffmpeg -hide_banner -v error -fflags nobuffer -flags low_delay -i http://192.168.68.151/flv?port=1935&app=bcs&stream=channel0_main.bcs&user=*&password=* -map 0:v:0? -c:v:0 copy -map 0:a:0? -c:a:0 copy -map 0:a:0? -c:a:1 libopus -application:a:1 lowdelay -min_comp 0 -user_agent ffmpeg/go2rtc -rtsp_transport tcp -f rtsp rtsp://127.0.0.1:8554/420517be3bed02c94ee20363840f2df3"
    },
    "555829": {
      "cpu": "0.0",
      "cpu_average": "0",
      "mem": "0.0",
      "cmdline": "sleep 30s"
    }
  },
  "service": {
    "uptime": 2455,
    "version": "0.17.0-f0d69f7",
    "latest_version": "0.17.0",
    "storage": {
      "/media/frigate/recordings": {
        "total": 1876684.2,
        "used": 1118778.2,
        "free": 662503.6,
        "mount_type": "ext4"
      },
      "/media/frigate/clips": {
        "total": 1876684.2,
        "used": 1118778.2,
        "free": 662503.6,
        "mount_type": "ext4"
      },
      "/tmp/cache": {
        "total": 953.7,
        "used": 24.3,
        "free": 929.4,
        "mount_type": "tmpfs"
      },
      "/dev/shm": {
        "total": 1024,
        "used": 488.8,
        "free": 535.2,
        "mount_type": "tmpfs",
        "available": 974,
        "camera_frame_size": 14.6,
        "shm_frame_count": 50,
        "min_shm": 342
      }
    },
    "temperatures": {

    },
    "last_updated": 1773993456
  },
  "processes": {
    "recording": {
      "pid": 802
    },
    "review_segment": {
      "pid": 815
    },
    "go2rtc": {
      "pid": 128
    },
    "embeddings": {
      "pid": 825
    }
  }
}

Operating system

Proxmox

Install method

Docker Compose

docker-compose file or Docker CLI command

services:

  frigate:
    container_name: frigate
    privileged: true
    restart: unless-stopped
    image: ghcr.io/blakeblackshear/frigate:0.17.0 #:stable
    cap_add:
      - CAP_PERFORM
    shm_size: "1024mb" # update for your cameras based on calculation above
    devices:
      - /dev/bus/usb:/dev/bus/usb
      - /dev/dri/renderD128 # for intel hwaccel, needs to be updated for your hardware
      - /dev/dri/card0 # for intel hwaccel, needs to be updated for your hardware
    volumes:
      - /etc/localtime:/etc/localtime:ro
      - /opt/frigate/config:/config
      - /mnt/frigate-data:/media/frigate
      - type: tmpfs # Optional: 1GB of memory, reduces SSD/SD Card wear
        target: /tmp/cache
        tmpfs:
          size: 1000000000
    ports:
      - "5000:5000"
      - "443:8971" #HTTPS
      - "1935:1935" # RTMP feeds
      - "8554:8554" # RTSP feeds
      - "8555:8555/tcp" # WebRTC over tcp
      - "8555:8555/udp" # WebRTC over udp
      - "1984:1984" #Go2Rtc interface
    environment:
      - PLUS_API_KEY=xxx
      - TZ=Europe/Paris
      - xxx keys
      - FFREPORT=file=/config/ffmpeg.log:level=32:format=text

Object Detector

Coral

Screenshots of the Frigate UI's System metrics pages

Any other information that may be helpful

No response

2026-03-20T08:10:30Z

inkeep-ai-assistant[bot]
bot Mar 20, 2026

Your image embedding time of ~800ms is within a reasonable range for CPU-based embedding, especially when using the small model size for semantic search as you have configured.(1) This is a separate process from object detection and does not directly cause false positives.(2)

Image embeddings are calculated when a tracked object's lifecycle ends, so the more detected objects you have, the more embeddings will be calculated and the more CPU is used.(2) The embedding speed metric reflects how long it takes to generate a vector representation of a thumbnail for semantic search and related features—not for real-time detection or recognition.(1)

False positives in detection are primarily influenced by your object detection model, thresholds, and filters—not by embedding speed.(3) If you are experiencing false positives, you should focus on:

Adjusting min_score and threshold values for your objects(1)
Submitting true and false positives to Frigate+ to improve your model(3)
Ensuring your training set is diverse and representative of your environment(3)

If you want to reduce embedding times, you could consider using a GPU for hardware acceleration, as the embedding model benefits significantly from a discrete GPU.(2)

📚 Sources:

Have another question? Just tag @inkeep.

0 replies

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[Config Support]: Understanding Image Embedding stats to see what i'm doing wrong #22544

Uh oh!

{{title}}

Uh oh!

Replies: 1 comment

Uh oh!

{{title}}

Uh oh!

Select a reply

Uh oh!

Uh oh!

[Config Support]: Understanding Image Embedding stats to see what i'm doing wrong #22544

Uh oh!

WorldOfGZ Mar 20, 2026

Describe the problem you are having

Version

Frigate config file

Relevant Frigate log output

Relevant go2rtc log output

Frigate stats

Operating system

Install method

docker-compose file or Docker CLI command

Object Detector

Screenshots of the Frigate UI's System metrics pages

Any other information that may be helpful

Replies: 1 comment

Uh oh!

inkeep-ai-assistant[bot] bot Mar 20, 2026

WorldOfGZ
Mar 20, 2026

inkeep-ai-assistant[bot]
bot Mar 20, 2026