Just installed 0.17.0 and it appears most object types are not configured #22162
pvanderwal
started this conversation in
General
Replies: 1 comment 1 reply
-
|
Those labels are not part of COCO, they are only available as part of Frigate+ models. https://docs.frigate.video/configuration/objects |
Beta Was this translation helpful? Give feedback.
1 reply
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
although they weren't supported under 0.16 either, since it looked like 0.17.0 was coming out soon I decided to wait.
So...running Frigate under docker on a NanoPi M5 (Rockchip RK3576)
When starting up I see the following:
WARNING : Goat_Run is configured to track ['raccoon', 'goat', 'squirrel', 'skunk'] objects, which are not supported by the current model.
and
WARNING : Goat_Shed is configured to track ['raccoon', 'goat', 'squirrel', 'skunk'] objects, which are not supported by the current model.
I see this on ever model I've tried (deci-fp16-yolonas_s, frigate-fp16-yolov9-t, rock-i8-yolox_nano, etc.)
This is my config.yaml:
tls:
enabled: false
mqtt:
enabled: true
host: mqtt.lan
topic_prefix: frigate
stats_interval: 60
ffmpeg:
hwaccel_args: preset-rkmpp
detectors:
rknn_0:
type: rknn
num_cores: 2
rknn_1:
type: rknn
num_cores: 2
record:
enabled: true
alerts:
retain:
days: 10
detections:
retain:
days: 10
model:
path: deci-fp16-yolonas_s
model_type: yolonas
input_pixel_format: bgr
path: frigate-fp16-yolov9-t
model_type: yolo-generic
width: 320
height: 320
path: rock-i8-yolox_nano
model_type: yolox
width: 416
height: 416
input_tensor: nhwc
labelmap_path: /labelmap/coco-80.txt
objects:
track:
- person
- dog
- cat
- bird
- goat
- squirrel
- raccoon
- skunk
- bear
filters:
person:
min_area: 5000
max_area: 100000
goat:
min_area: 500
max_area: 50000
dog:
min_area: 1000
max_area: 50000
go2rtc:
streams:
Goat_Shed:
- rtsp://Frigate:P@ssW0rd@192.168.128.250:554/stream1 # <- main camera stream
- hls # Enable HLS output for this stream
Goat_Run:
- rtsp://Frigate:P@ssW0rd@192.168.128.251:554/stream1 # <- main camera stream
- hls # Enable HLS output for this stream
webrtc:
candidates:
- stun:8555
cameras:
Goat_Run:
ffmpeg:
output_args:
record: preset-record-generic-audio-aac
inputs:
- path: rtsp://127.0.0.1:8554/Goat_Run # <--must match the name in go2rtc restream
input_args: preset-rtsp-restream
roles:
- record
- path: rtsp://Frigate:P@ssW0rd@192.168.128.251:554/stream2 #direct from camera
roles:
- detect
detect:
width: 640
height: 360
fps: 5
motion:
threshold: 30
contour_area: 35
improve_contrast: true
Goat_Shed:
ffmpeg:
output_args:
record: preset-record-generic-audio-aac
inputs:
- path: rtsp://127.0.0.1:8554/Goat_Shed # <--must match the name in go2rtc restream
input_args: preset-rtsp-restream
roles:
- record
- path: rtsp://Frigate:P@ssW0rd@192.168.128.250:554/stream2 #direct from camera
roles:
- detect
detect:
width: 640
height: 360
fps: 5
objects:
track:
- goat
- squirrel
- raccoon
- skunk
version: 0.17-0
camera_groups:
Goats:
order: 1
icon: LuDog
cameras:
- Goat_Run
- Goat_Shed
detect:
enabled: true
and this is my docker-compose.yml:
services:
frigate:
container_name: frigate
privileged: true
restart: unless-stopped
image: ghcr.io/blakeblackshear/frigate:stable-rk
image: ghcr.io/blakeblackshear/frigate:0.17.0-rc3-rk
services:
frigate:
container_name: frigate
privileged: true
restart: unless-stopped
image: ghcr.io/blakeblackshear/frigate:stable-rk
image: ghcr.io/blakeblackshear/frigate:0.17.0-rc3-rk
services:
frigate:
container_name: frigate
privileged: true
restart: unless-stopped
image: ghcr.io/blakeblackshear/frigate:stable-rk
image: ghcr.io/blakeblackshear/frigate:0.17.0-rc3-rk
services:
frigate:
container_name: frigate
privileged: true
restart: unless-stopped
image: ghcr.io/blakeblackshear/frigate:stable-rk
image: ghcr.io/blakeblackshear/frigate:0.17.0-rc3-rk
So are these object types just not included anymore, or did I do something wrong?
Beta Was this translation helpful? Give feedback.
All reactions