administrator/frigate
1
0
Fork 0
mirror of https://github.com/blakeblackshear/frigate.git synced 2026-03-23 16:48:23 +03:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

fix: add rk3588s support with model mapping to rk3588

Browse Source 
- Add rk3588s to SUPPORTED_RK_SOCS list
- Map rk3588s to rk3588 for model downloads (models are compatible)
- Update documentation to include rk3588s in supported SoCs
- Addresses feedback from blakeblackshear/frigate#20441
This commit is contained in:
Brian H 2026-02-01 17:38:21 +11:00
parent dea03ea099
commit a7042a7c0b
4 changed files with 7 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
docs/docs/configuration/object_detectors.md
View File

@ -1352,7 +1352,7 @@ To convert a onnx model to the rknn format using the [rknn-toolkit2](https://git
This is an example configuration file that you need to adjust to your specific onnx model: This is an example configuration file that you need to adjust to your specific onnx model:
```yaml ```yaml
soc: ["rk3562", "rk3566", "rk3568", "rk3576", "rk3588"] soc: ["rk3562", "rk3566", "rk3568", "rk3576", "rk3588", "rk3588s"]
quantization: false quantization: false
output_name: "{input_basename}" output_name: "{input_basename}"

1
docs/docs/frigate/hardware.md
View File

@ -273,6 +273,7 @@ Frigate supports hardware video processing on all Rockchip boards. However, hard
- RK3568 - RK3568
- RK3576 - RK3576
- RK3588 - RK3588
- RK3588S
| Name | YOLOv9 Inference Time | YOLO-NAS Inference Time | YOLOx Inference Time | | Name | YOLOv9 Inference Time | YOLO-NAS Inference Time | YOLOx Inference Time |
| -------------- | --------------------- | --------------------------- | ----------------------- | | -------------- | --------------------- | --------------------------- | ----------------------- |

2
frigate/const.py
View File

@ -78,7 +78,7 @@ FFMPEG_HWACCEL_AMF = "preset-amd-amf"
FFMPEG_HVC1_ARGS = ["-tag:v", "hvc1"] FFMPEG_HVC1_ARGS = ["-tag:v", "hvc1"]
# RKNN constants # RKNN constants
SUPPORTED_RK_SOCS = ["rk3562", "rk3566", "rk3568", "rk3576", "rk3588"] SUPPORTED_RK_SOCS = ["rk3562", "rk3566", "rk3568", "rk3576", "rk3588", "rk3588s"]
# Regex constants # Regex constants

5
frigate/detectors/plugins/rknn.py
View File

@ -91,6 +91,9 @@ class Rknn(DetectionApi):
def parse_model_input(self, model_path, soc): def parse_model_input(self, model_path, soc):
model_props = {} model_props = {}
# Map rk3588s to rk3588 for model compatibility
model_soc = "rk3588" if soc == "rk3588s" else soc
# find out if user provides his own model # find out if user provides his own model
# user provided models should be a path and contain a "/" # user provided models should be a path and contain a "/"
@ -148,7 +151,7 @@ class Rknn(DetectionApi):
model_props["model_type"] = model_type model_props["model_type"] = model_type
if model_matched: if model_matched:
model_props["filename"] = model_path + f"-{soc}-v2.3.2-2.rknn" model_props["filename"] = model_path + f"-{model_soc}-v2.3.2-2.rknn"
model_props["path"] = model_cache_dir + model_props["filename"] model_props["path"] = model_cache_dir + model_props["filename"]