administrator/frigate
1
0
Fork 0
mirror of https://github.com/blakeblackshear/frigate.git synced 2026-06-27 06:41:53 +03:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity
accbab7afc
frigate/docs/data/object_detectors_models.json

350 lines
71 KiB
JSON
Raw Normal View History

Rebuild object docs (#23570) * Rebuild object docs * Tweak styling and fix mixing tabs * Fix warning * Cleanup styling
2026-06-26 15:48:42 +03:00
{
"edgeTPU": {
"title": "EdgeTPU",
"models": [
{
"key": "mobiledet",
"label": "Mobiledet",
"recommended": true,
"download": "A TensorFlow Lite model is provided in the container at `/edgetpu_model.tflite` and is used by this detector type by default. To provide your own model, bind mount the file into the container and provide the path with `model.path`.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **EdgeTPU** from the detector type dropdown and click **Add**, then set device to `usb`.",
"yaml": "detectors:\n coral:\n type: edgetpu\n device: usb"
},
{
"key": "yolov9",
"label": "YOLOv9",
"recommended": false,
"download": "[Download the model](https://github.com/dbro/frigate-detector-edgetpu-yolo9/releases/download/v1.0/yolov9-s-relu6-best_320_int8_edgetpu.tflite), bind mount the file into the container, and provide the path with `model.path`. Note that the linked model requires a 17-label [labelmap file](https://raw.githubusercontent.com/dbro/frigate-detector-edgetpu-yolo9/refs/heads/main/labels-coco17.txt) that includes only 17 COCO classes.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **EdgeTPU** from the detector type dropdown and click **Add**, then set device to `usb`. Then on the same page, in the **Custom Model** tab, configure the model settings:\n\n| Field | Value |\n| ---------------------------------------- | ----------------------------------------------------------------- |\n| **Object Detection Model Type** | `yolo-generic` |\n| **Object detection model input width** | `320` (should match the imgsize of the model) |\n| **Object detection model input height** | `320` (should match the imgsize of the model) |\n| **Custom object detector model path** | `/config/model_cache/yolov9-s-relu6-best_320_int8_edgetpu.tflite` |\n| **Label map for custom object detector** | `/config/labels-coco17.txt` |",
"yaml": "detectors:\n coral:\n type: edgetpu\n device: usb\n\nmodel:\n model_type: yolo-generic\n width: 320 # <--- should match the imgsize of the model, typically 320\n height: 320 # <--- should match the imgsize of the model, typically 320\n path: /config/model_cache/yolov9-s-relu6-best_320_int8_edgetpu.tflite\n labelmap_path: /config/labels-coco17.txt"
}
]
},
"hailo8l": {
"title": "Hailo-8/Hailo-8L",
"models": [
{
"key": "yolo",
"label": "YOLO",
"recommended": true,
"download": "If no custom model path or URL is provided, the Hailo detector automatically downloads the default model (YOLOv6n) from the Hailo Model Zoo on first startup based on the detected hardware. Once cached under `/config/model_cache/hailo`, the model works fully offline.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **Hailo-8/Hailo-8L** from the detector type dropdown and click **Add**, then set device to `PCIe`. Then on the same page, in the **Custom Model** tab, configure the model settings:\n\n| Field | Value |\n| ---------------------------------------- | ----------------------- |\n| **Object detection model input width** | `320` |\n| **Object detection model input height** | `320` |\n| **Model Input Tensor Shape** | `nhwc` |\n| **Model Input Pixel Color Format** | `rgb` |\n| **Model Input D Type** | `int` |\n| **Object Detection Model Type** | `yolo-generic` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |\n\nThe detector automatically selects the default model based on your hardware. Optionally, specify a local model path or URL to override.",
"yaml": "detectors:\n hailo:\n type: hailo8l\n device: PCIe\n\nmodel:\n width: 320\n height: 320\n input_tensor: nhwc\n input_pixel_format: rgb\n input_dtype: int\n model_type: yolo-generic\n labelmap_path: /labelmap/coco-80.txt\n\n # The detector automatically selects the default model based on your hardware:\n # - For Hailo-8 hardware: YOLOv6n (default: yolov6n.hef)\n # - For Hailo-8L hardware: YOLOv6n (default: yolov6n.hef)\n #\n # Optionally, you can specify a local model path to override the default.\n # If a local path is provided and the file exists, it will be used instead of downloading.\n # Example:\n # path: /config/model_cache/hailo/yolov6n.hef\n #\n # You can also override using a custom URL:\n # path: https://hailo-model-zoo.s3.eu-west-2.amazonaws.com/ModelZoo/Compiled/v2.14.0/hailo8/yolov6n.hef\n # just make sure to give it the write configuration based on the model"
},
{
"key": "ssd",
"label": "SSD MobileNet v1",
"recommended": false,
"download": "For SSD-based models, provide either a model path or URL to your compiled SSD model. The integration will first check the local path before downloading if necessary. The model file is cached under `/config/model_cache/hailo`.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **Hailo-8/Hailo-8L** from the detector type dropdown and click **Add**, then set device to `PCIe`. Then on the same page, in the **Custom Model** tab, configure the model settings:\n\n| Field | Value |\n| --------------------------------------- | ------ |\n| **Object detection model input width** | `300` |\n| **Object detection model input height** | `300` |\n| **Model Input Tensor Shape** | `nhwc` |\n| **Model Input Pixel Color Format** | `rgb` |\n| **Object Detection Model Type** | `ssd` |\n\nSpecify the local model path or URL for SSD MobileNet v1.",
"yaml": "detectors:\n hailo:\n type: hailo8l\n device: PCIe\n\nmodel:\n width: 300\n height: 300\n input_tensor: nhwc\n input_pixel_format: rgb\n model_type: ssd\n # Specify the local model path (if available) or URL for SSD MobileNet v1.\n # Example with a local path:\n # path: /config/model_cache/h8l_cache/ssd_mobilenet_v1.hef\n #\n # Or override using a custom URL:\n # path: https://hailo-model-zoo.s3.eu-west-2.amazonaws.com/ModelZoo/Compiled/v2.14.0/hailo8l/ssd_mobilenet_v1.hef"
}
]
},
"openvino": {
"title": "OpenVINO",
"models": [
{
"key": "ssd",
"label": "SSDLite MobileNet v2",
"recommended": true,
"download": "An OpenVINO model is provided in the container at `/openvino-model/ssdlite_mobilenet_v2.xml` and is used by this detector type by default. The model comes from Intel's Open Model Zoo [SSDLite MobileNet V2](https://github.com/openvinotoolkit/open_model_zoo/tree/master/models/public/ssdlite_mobilenet_v2) and is converted to an FP16 precision IR model.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **OpenVINO** from the detector type dropdown and click **Add**, then set device to `GPU` (or `NPU`). Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | ------------------------------------------ |\n| **Object detection model input width** | `300` |\n| **Object detection model input height** | `300` |\n| **Model Input Tensor Shape** | `nhwc` |\n| **Model Input Pixel Color Format** | `bgr` |\n| **Custom object detector model path** | `/openvino-model/ssdlite_mobilenet_v2.xml` |\n| **Label map for custom object detector** | `/openvino-model/coco_91cl_bkgr.txt` |",
"yaml": "detectors:\n ov:\n type: openvino\n device: GPU # Or NPU\n\nmodel:\n width: 300\n height: 300\n input_tensor: nhwc\n input_pixel_format: bgr\n path: /openvino-model/ssdlite_mobilenet_v2.xml\n labelmap_path: /openvino-model/coco_91cl_bkgr.txt"
},
{
"key": "yolov9",
"label": "YOLOv9",
"recommended": false,
"download": "YOLOv9 model can be exported as ONNX using the command below. You can copy and paste the whole thing to your terminal and execute, altering `MODEL_SIZE=t` and `IMG_SIZE=320` in the first line to the [model size](https://github.com/WongKinYiu/yolov9#performance) you would like to convert (available model sizes are `t`, `s`, `m`, `c`, and `e`, common image sizes are `320` and `640`).\n\n```sh\ndocker build . --build-arg MODEL_SIZE=t --build-arg IMG_SIZE=320 --output . -f- <<'EOF'\nFROM python:3.11 AS build\nRUN apt-get update && apt-get install --no-install-recommends -y cmake libgl1 && rm -rf /var/lib/apt/lists/*\nCOPY --from=ghcr.io/astral-sh/uv:0.10.4 /uv /bin/\nWORKDIR /yolov9\nADD https://github.com/WongKinYiu/yolov9.git .\nRUN uv pip install --system -r requirements.txt\nRUN uv pip install --system onnx==1.18.0 onnxruntime onnx-simplifier==0.4.* onnxscript\nARG MODEL_SIZE\nARG IMG_SIZE\nADD https://github.com/WongKinYiu/yolov9/releases/download/v0.1/yolov9-${MODEL_SIZE}-converted.pt yolov9-${MODEL_SIZE}.pt\nRUN sed -i \"s/ckpt = torch.load(attempt_download(w), map_location='cpu')/ckpt = torch.load(attempt_download(w), map_location='cpu', weights_only=False)/g\" models/experimental.py\nRUN python3 export.py --weights ./yolov9-${MODEL_SIZE}.pt --imgsz ${IMG_SIZE} --simplify --include onnx\nFROM scratch\nARG MODEL_SIZE\nARG IMG_SIZE\nCOPY --from=build /yolov9/yolov9-${MODEL_SIZE}.onnx /yolov9-${MODEL_SIZE}-${IMG_SIZE}.onnx\nEOF\n```",
"ui": "Navigate to **Settings > System > Detectors and model** and select **OpenVINO** from the detector type dropdown and click **Add**, then set device to `GPU` (or `NPU`). Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | -------------------------------------------------------- |\n| **Object Detection Model Type** | `yolo-generic` |\n| **Object detection model input width** | `320` (should match the imgsize set during model export) |\n| **Object detection model input height** | `320` (should match the imgsize set during model export) |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input D Type** | `float` |\n| **Custom object detector model path** | `/config/model_cache/yolo.onnx` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors:\n ov:\n type: openvino\n device: GPU # or NPU\n\nmodel:\n model_type: yolo-generic\n width: 320 # <--- should match the imgsize set during model export\n height: 320 # <--- should match the imgsize set during model export\n input_tensor: nchw\n input_dtype: float\n path: /config/model_cache/yolo.onnx\n labelmap_path: /labelmap/coco-80.txt"
},
{
"key": "yolo-legacy",
"label": "YOLO (v3, v4, v7)",
"recommended": false,
"download": "To export as ONNX:\n\n```sh\ngit clone https://github.com/NateMeyer/tensorrt_demos\ncd tensorrt_demos/yolo\n./download_yolo.sh\npython3 yolo_to_onnx.py -m yolov7-320\n```",
"ui": "Navigate to **Settings > System > Detectors and model** and select **OpenVINO** from the detector type dropdown and click **Add**, then set device to `GPU` (or `NPU`). Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | -------------------------------------------------------- |\n| **Object Detection Model Type** | `yolo-generic` |\n| **Object detection model input width** | `320` (should match the imgsize set during model export) |\n| **Object detection model input height** | `320` (should match the imgsize set during model export) |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input D Type** | `float` |\n| **Custom object detector model path** | `/config/model_cache/yolo.onnx` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors:\n ov:\n type: openvino\n device: GPU # or NPU\n\nmodel:\n model_type: yolo-generic\n width: 320 # <--- should match the imgsize set during model export\n height: 320 # <--- should match the imgsize set during model export\n input_tensor: nchw\n input_dtype: float\n path: /config/model_cache/yolo.onnx\n labelmap_path: /labelmap/coco-80.txt"
},
{
"key": "yolonas",
"label": "YOLO-NAS",
"recommended": false,
"download": "You can build and download a compatible model with pre-trained weights using [this notebook](https://github.com/blakeblackshear/frigate/blob/dev/notebooks/YOLO_NAS_Pretrained_Export.ipynb) [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/blakeblackshear/frigate/blob/dev/notebooks/YOLO_NAS_Pretrained_Export.ipynb) which can be run directly in [Google Colab](https://colab.research.google.com/github/blakeblackshear/frigate/blob/dev/notebooks/YOLO_NAS_Pretrained_Export.ipynb).\n\n:::warning\n\nThe pre-trained YOLO-NAS weights from DeciAI are subject to their license and can't be used commercially. For more information, see: https://docs.deci.ai/super-gradients/latest/LICENSE.YOLONAS.html\n\n:::\n\nThe input image size in this notebook is set to 320x320. This results in lower CPU usage and faster inference times without impacting performance in most cases due to the way Frigate crops video frames to areas of interest before running detection. The notebook and config can be updated to 640x640 if desired.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **OpenVINO** from the detector type dropdown and click **Add**, then set device to `GPU`. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | ------------------------------------------------- |\n| **Object Detection Model Type** | `yolonas` |\n| **Object detection model input width** | `320` (should match whatever was set in notebook) |\n| **Object detection model input height** | `320` (should match whatever was set in notebook) |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input Pixel Color Format** | `bgr` |\n| **Custom object detector model path** | `/config/yolo_nas_s.onnx` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors:\n ov:\n type: openvino\n device: GPU\n\nmodel:\n model_type: yolonas\n width: 320 # <--- should match whatever was set in notebook\n height: 320 # <--- should match whatever was set in notebook\n input_tensor: nchw\n input_pixel_format: bgr\n path: /config/yolo_nas_s.onnx\n labelmap_path: /labelmap/coco-80.txt"
},
{
"key": "yolox",
"label": "YOLOX",
"recommended": false,
"download": "YOLOx models can be downloaded [from the YOLOx repo](https://github.com/Megvii-BaseDetection/YOLOX/tree/main/demo/ONNXRuntime).",
"ui": "Navigate to **Settings > System > Detectors and model** and select **OpenVINO** from the detector type dropdown and click **Add**, then set device to `GPU`. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ------------------------------------- | -------------------------------- |\n| **Object Detection Model Type** | `yolox` |\n| **Custom object detector model path** | path to your YOLOX ONNX model |",
"yaml": "detectors:\n ov:\n type: openvino\n device: GPU\n\nmodel:\n model_type: yolox\n path: /config/model_cache/yolox.onnx\n labelmap_path: /labelmap/coco-80.txt"
},
{
"key": "rfdetr",
"label": "RF-DETR",
"recommended": false,
"download": "RF-DETR can be exported as ONNX by running the command below. You can copy and paste the whole thing to your terminal and execute, altering `MODEL_SIZE=Nano` in the first line to `Nano`, `Small`, or `Medium` size.\n\n```sh\ndocker build . --build-arg MODEL_SIZE=Nano --rm --output . -f- <<'EOF'\nFROM python:3.12 AS build\nRUN apt-get update && apt-get install --no-install-recommends -y libgl1 && rm -rf /var/lib/apt/lists/*\nCOPY --from=ghcr.io/astral-sh/uv:0.10.4 /uv /bin/\nWORKDIR /rfdetr\nRUN uv pip install --system rfdetr[onnxexport] torch==2.8.0 onnx==1.19.1 transformers==4.57.6 onnxscript\nARG MODEL_SIZE\nRUN python3 -c \"from rfdetr import RFDETR${MODEL_SIZE}; x = RFDETR${MODEL_SIZE}(resolution=320); x.export(simplify=True)\"\nFROM scratch\nARG MODEL_SIZE\nCOPY --from=build /rfdetr/output/inference_model.onnx /rfdetr-${MODEL_SIZE}.onnx\nEOF\n```",
"ui": "Navigate to **Settings > System > Detectors and model** and select **OpenVINO** from the detector type dropdown and click **Add**, then set device to `GPU`. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| --------------------------------------- | --------------------------------- |\n| **Object Detection Model Type** | `rfdetr` |\n| **Object detection model input width** | `320` |\n| **Object detection model input height** | `320` |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input D Type** | `float` |\n| **Custom object detector model path** | `/config/model_cache/rfdetr.onnx` |",
"yaml": "detectors:\n ov:\n type: openvino\n device: GPU\n\nmodel:\n model_type: rfdetr\n width: 320\n height: 320\n input_tensor: nchw\n input_dtype: float\n path: /config/model_cache/rfdetr.onnx"
},
{
"key": "dfine",
"label": "D-FINE / DEIMv2",
"recommended": false,
"download": "#### D-FINE\n\nD-FINE can be exported as ONNX by running the command below. You can copy and paste the whole thing to your terminal and execute, altering `MODEL_SIZE=s` in the first line to `s`, `m`, or `l` size.\n\n```sh\ndocker build . --build-arg MODEL_SIZE=s --output . -f- <<'EOF'\nFROM python:3.11 AS build\nRUN apt-get update && apt-get install --no-install-recommends -y libgl1 && rm -rf /var/lib/apt/lists/*\nCOPY --from=ghcr.io/astral-sh/uv:0.8.0 /uv /bin/\nWORKDIR /dfine\nRUN git clone https://github.com/Peterande/D-FINE.git .\nRUN uv pip install --system -r requirements.txt\nRUN uv pip install --system onnx onnxruntime onnxsim onnxscript\n# Create output directory and download checkpoint\nRUN mkdir -p output\nARG MODEL_SIZE\nRUN wget https://github.com/Peterande/storage/releases/download/dfinev1.0/dfine_${MODEL_SIZE}_obj2coco.pth -O output/dfine_${MODEL_SIZE}_obj2coco.pth\n# Modify line 58 of export_onnx.py to change batch size to 1\nRUN sed -i '58s/data = torch.rand(.*)/data = torch.rand(1, 3, 640, 640)/' tools/deployment/export_onnx.py\nRUN python3 tools/deployment/export_onnx.py -c configs/dfine/objects365/dfine_hgnetv2_${MODEL_SIZE}_obj2coco.yml -r output/dfine_${MODEL_SIZE}_obj2coco.pth\nFROM scratch\nARG MODEL_SIZE\nCOPY --from=build /dfine/output/dfine_${MODEL_SIZE}_obj2coco.onnx /dfine-${MODEL_SIZE}.onnx\nEOF\n```\n\n#### DEIMv2\n\n[DEIMv2](https://github.com/Intellindust-AI-Lab/DEIMv2) can be exported as ONNX by running the command below. Pretrained weights are available on Hugging Face for two backbone families:\n\n- **HGNetv2** (smaller/faster): `atto`, `femto`, `pico`, `n`\n- **DINOv3** (larger/more accurate): `s`, `m`, `l`, `x`\n\nSet `BACKBONE` and `MODEL_SIZE` in the first line to match your desired variant. Hugging Face model names use uppercase (e.g. `HGNetv2_N`, `DINOv3_S`), while config files use lowercase (e.g. `hgnetv2_n`, `dinov3_s`).\n\n```sh\ndocker build . --rm --build-arg BACKBONE=hgnetv2 --build-arg MODEL_SIZE=n --output . -f- <<'EOF'\nFROM python:3.11-slim AS build\nRUN apt-get update && apt-get install --no-install-recommends -y git libgl1 libglib2.0-0 && rm -rf /var/lib/apt/lists/*\nCOPY --from=ghcr.io/astral-sh/uv:0.8.0 /uv /bin/\nWORKDIR /deimv2\nRUN git clone https://github.com/Intellindust-AI-Lab/DEIMv2.git .\n# Install CPU-only PyTorch first to avoid pulling CUDA variant\nRUN uv pip install --no-cache --system torch torchvision --index-url https://download.pytorch.org/whl/cpu\nRUN uv pip install --no-cache --system -r requirements.txt\nRUN uv pip install --no-cache --system onnx safetensors huggingface_hub\nRUN mkdir -p output\nARG BACKBONE\nARG MODEL_SIZE\n# Download from Hugging Face and convert safetensors to pth\nRUN python3 -c \"\\\nfrom huggingface_hub import hf_hub_download; \\\nfrom safetensors.torch import load_file; \\\nimport torch; \\\nbackbone = '${BACKBONE}'.replace('hgnetv2','HGNetv2').replace('dinov3','DINOv3'); \\\nsize = '${MODEL_SIZE}'.upper(); \\\nst = load_file(hf_hub_download('Intellindust/DEIMv2_' + backbone + '_' + size + '_COCO', 'model.safetensors')); \\\ntorch.save({'model': st}, 'output/deimv2.pth')\"\nRUN sed -i \"s/data = torch.rand(2/data = torch.rand(1/\" tools/deployment/export_onnx.py\n# HuggingFace safetensors omits frozen constants that the model constructor initializes\nRUN sed -i \"s/cfg.model.load_state_dict(state)/cfg.model.load_state_dict(state, strict=False)/\" tools/deployment/export_onnx.py\nRUN python3 tools/deployment/export_onnx.py -c configs/deimv2/deimv2_${BACKBONE}_${MODEL_SIZE}_coco.yml -r output/deimv2.pth\nFROM scratch\nARG BACKBONE\nARG MODEL_SIZE\nCOPY --from=build /deimv2/output/deimv2.onnx /deimv2_${BACKBONE}_${MODEL_SIZE}.onnx\nEOF\n```",
"ui": "Navigate to **Settings > System > Detectors and model** and select **OpenVINO** from the detector type dropdown and click **Add**, then set device to `CPU`. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | ---------------------------------- |\n| **Object Detection Model Type** | `dfine` |\n| **Object detection model input width** | `640` |\n| **Object detection model input height** | `640` |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input D Type** | `float` |\n| **Custom object detector model path** | `/config/model_cache/dfine-s.onnx` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors:\n ov:\n type: openvino\n device: CPU\n\nmodel:\n model_type: dfine\n width: 640\n height: 640\n input_tensor: nchw\n input_dtype: float\n path: /config/model_cache/dfine-s.onnx\n labelmap_path: /labelmap/coco-80.txt"
}
]
},
"appleSilicon": {
"title": "Apple Silicon",
"models": [
{
"key": "yolov9",
"label": "YOLOv9",
"recommended": true,
"download": "YOLOv9 model can be exported as ONNX using the command below. You can copy and paste the whole thing to your terminal and execute, altering `MODEL_SIZE=t` and `IMG_SIZE=320` in the first line to the [model size](https://github.com/WongKinYiu/yolov9#performance) you would like to convert (available model sizes are `t`, `s`, `m`, `c`, and `e`, common image sizes are `320` and `640`).\n\n```sh\ndocker build . --build-arg MODEL_SIZE=t --build-arg IMG_SIZE=320 --output . -f- <<'EOF'\nFROM python:3.11 AS build\nRUN apt-get update && apt-get install --no-install-recommends -y cmake libgl1 && rm -rf /var/lib/apt/lists/*\nCOPY --from=ghcr.io/astral-sh/uv:0.10.4 /uv /bin/\nWORKDIR /yolov9\nADD https://github.com/WongKinYiu/yolov9.git .\nRUN uv pip install --system -r requirements.txt\nRUN uv pip install --system onnx==1.18.0 onnxruntime onnx-simplifier==0.4.* onnxscript\nARG MODEL_SIZE\nARG IMG_SIZE\nADD https://github.com/WongKinYiu/yolov9/releases/download/v0.1/yolov9-${MODEL_SIZE}-converted.pt yolov9-${MODEL_SIZE}.pt\nRUN sed -i \"s/ckpt = torch.load(attempt_download(w), map_location='cpu')/ckpt = torch.load(attempt_download(w), map_location='cpu', weights_only=False)/g\" models/experimental.py\nRUN python3 export.py --weights ./yolov9-${MODEL_SIZE}.pt --imgsz ${IMG_SIZE} --simplify --include onnx\nFROM scratch\nARG MODEL_SIZE\nARG IMG_SIZE\nCOPY --from=build /yolov9/yolov9-${MODEL_SIZE}.onnx /yolov9-${MODEL_SIZE}-${IMG_SIZE}.onnx\nEOF\n```",
"ui": "Navigate to **Settings > System > Detectors and model** and select **ZMQ IPC** from the detector type dropdown and click **Add**, then set the endpoint to `tcp://host.docker.internal:5555`. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | -------------------------------------------------------- |\n| **Object Detection Model Type** | `yolo-generic` |\n| **Object detection model input width** | `320` (should match the imgsize set during model export) |\n| **Object detection model input height** | `320` (should match the imgsize set during model export) |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input D Type** | `float` |\n| **Custom object detector model path** | `/config/model_cache/yolo.onnx` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors:\n apple-silicon:\n type: zmq\n endpoint: tcp://host.docker.internal:5555\n\nmodel:\n model_type: yolo-generic\n width: 320 # <--- should match the imgsize set during model export\n height: 320 # <--- should match the imgsize set during model export\n input_tensor: nchw\n input_dtype: float\n path: /config/model_cache/yolo.onnx\n labelmap_path: /labelmap/coco-80.txt"
},
{
"key": "yolo-legacy",
"label": "YOLO (v3, v4, v7)",
"recommended": false,
"download": "To export as ONNX:\n\n```sh\ngit clone https://github.com/NateMeyer/tensorrt_demos\ncd tensorrt_demos/yolo\n./download_yolo.sh\npython3 yolo_to_onnx.py -m yolov7-320\n```",
"ui": "Navigate to **Settings > System > Detectors and model** and select **ZMQ IPC** from the detector type dropdown and click **Add**, then set the endpoint to `tcp://host.docker.internal:5555`. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | -------------------------------------------------------- |\n| **Object Detection Model Type** | `yolo-generic` |\n| **Object detection model input width** | `320` (should match the imgsize set during model export) |\n| **Object detection model input height** | `320` (should match the imgsize set during model export) |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input D Type** | `float` |\n| **Custom object detector model path** | `/config/model_cache/yolo.onnx` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors:\n apple-silicon:\n type: zmq\n endpoint: tcp://host.docker.internal:5555\n\nmodel:\n model_type: yolo-generic\n width: 320 # <--- should match the imgsize set during model export\n height: 320 # <--- should match the imgsize set during model export\n input_tensor: nchw\n input_dtype: float\n path: /config/model_cache/yolo.onnx\n labelmap_path: /labelmap/coco-80.txt"
}
]
},
"onnx": {
"title": "ONNX",
"models": [
{
"key": "yolov9",
"label": "YOLOv9",
"recommended": true,
"download": "YOLOv9 model can be exported as ONNX using the command below. You can copy and paste the whole thing to your terminal and execute, altering `MODEL_SIZE=t` and `IMG_SIZE=320` in the first line to the [model size](https://github.com/WongKinYiu/yolov9#performance) you would like to convert (available model sizes are `t`, `s`, `m`, `c`, and `e`, common image sizes are `320` and `640`).\n\n```sh\ndocker build . --build-arg MODEL_SIZE=t --build-arg IMG_SIZE=320 --output . -f- <<'EOF'\nFROM python:3.11 AS build\nRUN apt-get update && apt-get install --no-install-recommends -y cmake libgl1 && rm -rf /var/lib/apt/lists/*\nCOPY --from=ghcr.io/astral-sh/uv:0.10.4 /uv /bin/\nWORKDIR /yolov9\nADD https://github.com/WongKinYiu/yolov9.git .\nRUN uv pip install --system -r requirements.txt\nRUN uv pip install --system onnx==1.18.0 onnxruntime onnx-simplifier==0.4.* onnxscript\nARG MODEL_SIZE\nARG IMG_SIZE\nADD https://github.com/WongKinYiu/yolov9/releases/download/v0.1/yolov9-${MODEL_SIZE}-converted.pt yolov9-${MODEL_SIZE}.pt\nRUN sed -i \"s/ckpt = torch.load(attempt_download(w), map_location='cpu')/ckpt = torch.load(attempt_download(w), map_location='cpu', weights_only=False)/g\" models/experimental.py\nRUN python3 export.py --weights ./yolov9-${MODEL_SIZE}.pt --imgsz ${IMG_SIZE} --simplify --include onnx\nFROM scratch\nARG MODEL_SIZE\nARG IMG_SIZE\nCOPY --from=build /yolov9/yolov9-${MODEL_SIZE}.onnx /yolov9-${MODEL_SIZE}-${IMG_SIZE}.onnx\nEOF\n```",
"ui": "Navigate to **Settings > System > Detectors and model** and select **ONNX** from the detector type dropdown and click **Add**. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | -------------------------------------------------------- |\n| **Object Detection Model Type** | `yolo-generic` |\n| **Object detection model input width** | `320` (should match the imgsize set during model export) |\n| **Object detection model input height** | `320` (should match the imgsize set during model export) |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input D Type** | `float` |\n| **Custom object detector model path** | `/config/model_cache/yolo.onnx` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors:\n onnx:\n type: onnx\n\nmodel:\n model_type: yolo-generic\n width: 320 # <--- should match the imgsize set during model export\n height: 320 # <--- should match the imgsize set during model export\n input_tensor: nchw\n input_dtype: float\n path: /config/model_cache/yolo.onnx\n labelmap_path: /labelmap/coco-80.txt"
},
{
"key": "rfdetr",
"label": "RF-DETR",
"recommended": false,
"download": "RF-DETR can be exported as ONNX by running the command below. You can copy and paste the whole thing to your terminal and execute, altering `MODEL_SIZE=Nano` in the first line to `Nano`, `Small`, or `Medium` size.\n\n```sh\ndocker build . --build-arg MODEL_SIZE=Nano --rm --output . -f- <<'EOF'\nFROM python:3.12 AS build\nRUN apt-get update && apt-get install --no-install-recommends -y libgl1 && rm -rf /var/lib/apt/lists/*\nCOPY --from=ghcr.io/astral-sh/uv:0.10.4 /uv /bin/\nWORKDIR /rfdetr\nRUN uv pip install --system rfdetr[onnxexport] torch==2.8.0 onnx==1.19.1 transformers==4.57.6 onnxscript\nARG MODEL_SIZE\nRUN python3 -c \"from rfdetr import RFDETR${MODEL_SIZE}; x = RFDETR${MODEL_SIZE}(resolution=320); x.export(simplify=True)\"\nFROM scratch\nARG MODEL_SIZE\nCOPY --from=build /rfdetr/output/inference_model.onnx /rfdetr-${MODEL_SIZE}.onnx\nEOF\n```",
"ui": "Navigate to **Settings > System > Detectors and model** and select **ONNX** from the detector type dropdown and click **Add**. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| --------------------------------------- | --------------------------------- |\n| **Object Detection Model Type** | `rfdetr` |\n| **Object detection model input width** | `320` |\n| **Object detection model input height** | `320` |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input D Type** | `float` |\n| **Custom object detector model path** | `/config/model_cache/rfdetr.onnx` |",
"yaml": "detectors:\n onnx:\n type: onnx\n\nmodel:\n model_type: rfdetr\n width: 320\n height: 320\n input_tensor: nchw\n input_dtype: float\n path: /config/model_cache/rfdetr.onnx"
},
{
"key": "yolonas",
"label": "YOLO-NAS",
"recommended": false,
"download": "You can build and download a compatible model with pre-trained weights using [this notebook](https://github.com/blakeblackshear/frigate/blob/dev/notebooks/YOLO_NAS_Pretrained_Export.ipynb) [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/blakeblackshear/frigate/blob/dev/notebooks/YOLO_NAS_Pretrained_Export.ipynb) which can be run directly in [Google Colab](https://colab.research.google.com/github/blakeblackshear/frigate/blob/dev/notebooks/YOLO_NAS_Pretrained_Export.ipynb).\n\n:::warning\n\nThe pre-trained YOLO-NAS weights from DeciAI are subject to their license and can't be used commercially. For more information, see: https://docs.deci.ai/super-gradients/latest/LICENSE.YOLONAS.html\n\n:::\n\nThe input image size in this notebook is set to 320x320. This results in lower CPU usage and faster inference times without impacting performance in most cases due to the way Frigate crops video frames to areas of interest before running detection. The notebook and config can be updated to 640x640 if desired.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **ONNX** from the detector type dropdown and click **Add**. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | ------------------------------------------------- |\n| **Object Detection Model Type** | `yolonas` |\n| **Object detection model input width** | `320` (should match whatever was set in notebook) |\n| **Object detection model input height** | `320` (should match whatever was set in notebook) |\n| **Model Input Pixel Color Format** | `bgr` |\n| **Model Input Tensor Shape** | `nchw` |\n| **Custom object detector model path** | `/config/yolo_nas_s.onnx` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors:\n onnx:\n type: onnx\n\nmodel:\n model_type: yolonas\n width: 320 # <--- should match whatever was set in notebook\n height: 320 # <--- should match whatever was set in notebook\n input_pixel_format: bgr\n input_tensor: nchw\n path: /config/yolo_nas_s.onnx\n labelmap_path: /labelmap/coco-80.txt"
},
{
"key": "yolox",
"label": "YOLOX",
"recommended": false,
"download": "YOLOx models can be downloaded [from the YOLOx repo](https://github.com/Megvii-BaseDetection/YOLOX/tree/main/demo/ONNXRuntime).",
"ui": "Navigate to **Settings > System > Detectors and model** and select **ONNX** from the detector type dropdown and click **Add**. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | -------------------------------------------------------- |\n| **Object Detection Model Type** | `yolox` |\n| **Object detection model input width** | `416` (should match the imgsize set during model export) |\n| **Object detection model input height** | `416` (should match the imgsize set during model export) |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input D Type** | `float_denorm` |\n| **Custom object detector model path** | `/config/model_cache/yolox_tiny.onnx` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors:\n onnx:\n type: onnx\n\nmodel:\n model_type: yolox\n width: 416 # <--- should match the imgsize set during model export\n height: 416 # <--- should match the imgsize set during model export\n input_tensor: nchw\n input_dtype: float_denorm\n path: /config/model_cache/yolox_tiny.onnx\n labelmap_path: /labelmap/coco-80.txt"
},
{
"key": "dfine",
"label": "D-FINE / DEIMv2",
"recommended": false,
"download": "#### Downloading D-FINE Model\n\nD-FINE can be exported as ONNX by running the command below. You can copy and paste the whole thing to your terminal and execute, altering `MODEL_SIZE=s` in the first line to `s`, `m`, or `l` size.\n\n```sh\ndocker build . --build-arg MODEL_SIZE=s --output . -f- <<'EOF'\nFROM python:3.11 AS build\nRUN apt-get update && apt-get install --no-install-recommends -y libgl1 && rm -rf /var/lib/apt/lists/*\nCOPY --from=ghcr.io/astral-sh/uv:0.8.0 /uv /bin/\nWORKDIR /dfine\nRUN git clone https://github.com/Peterande/D-FINE.git .\nRUN uv pip install --system -r requirements.txt\nRUN uv pip install --system onnx onnxruntime onnxsim onnxscript\n# Create output directory and download checkpoint\nRUN mkdir -p output\nARG MODEL_SIZE\nRUN wget https://github.com/Peterande/storage/releases/download/dfinev1.0/dfine_${MODEL_SIZE}_obj2coco.pth -O output/dfine_${MODEL_SIZE}_obj2coco.pth\n# Modify line 58 of export_onnx.py to change batch size to 1\nRUN sed -i '58s/data = torch.rand(.*)/data = torch.rand(1, 3, 640, 640)/' tools/deployment/export_onnx.py\nRUN python3 tools/deployment/export_onnx.py -c configs/dfine/objects365/dfine_hgnetv2_${MODEL_SIZE}_obj2coco.yml -r output/dfine_${MODEL_SIZE}_obj2coco.pth\nFROM scratch\nARG MODEL_SIZE\nCOPY --from=build /dfine/output/dfine_${MODEL_SIZE}_obj2coco.onnx /dfine-${MODEL_SIZE}.onnx\nEOF\n```\n\n#### Downloading DEIMv2 Model\n\n[DEIMv2](https://github.com/Intellindust-AI-Lab/DEIMv2) can be exported as ONNX by running the command below. Pretrained weights are available on Hugging Face for two backbone families:\n\n- **HGNetv2** (smaller/faster): `atto`, `femto`, `pico`, `n`\n- **DINOv3** (larger/more accurate): `s`, `m`, `l`, `x`\n\nSet `BACKBONE` and `MODEL_SIZE` in the first line to match your desired variant. Hugging Face model names use uppercase (e.g. `HGNetv2_N`, `DINOv3_S`), while config files use lowercase (e.g. `hgnetv2_n`, `dinov3_s`).\n\n```sh\ndocker build . --rm --build-arg BACKBONE=hgnetv2 --build-arg MODEL_SIZE=n --output . -f- <<'EOF'\nFROM python:3.11-slim AS build\nRUN apt-get update && apt-get install --no-install-recommends -y git libgl1 libglib2.0-0 && rm -rf /var/lib/apt/lists/*\nCOPY --from=ghcr.io/astral-sh/uv:0.8.0 /uv /bin/\nWORKDIR /deimv2\nRUN git clone https://github.com/Intellindust-AI-Lab/DEIMv2.git .\n# Install CPU-only PyTorch first to avoid pulling CUDA variant\nRUN uv pip install --no-cache --system torch torchvision --index-url https://download.pytorch.org/whl/cpu\nRUN uv pip install --no-cache --system -r requirements.txt\nRUN uv pip install --no-cache --system onnx safetensors huggingface_hub\nRUN mkdir -p output\nARG BACKBONE\nARG MODEL_SIZE\n# Download from Hugging Face and convert safetensors to pth\nRUN python3 -c \"\\\nfrom huggingface_hub import hf_hub_download; \\\nfrom safetensors.torch import load_file; \\\nimport torch; \\\nbackbone = '${BACKBONE}'.replace('hgnetv2','HGNetv2').replace('dinov3','DINOv3'); \\\nsize = '${MODEL_SIZE}'.upper(); \\\nst = load_file(hf_hub_download('Intellindust/DEIMv2_' + backbone + '_' + size + '_COCO', 'model.safetensors')); \\\ntorch.save({'model': st}, 'output/deimv2.pth')\"\nRUN sed -i \"s/data = torch.rand(2/data = torch.rand(1/\" tools/deployment/export_onnx.py\n# HuggingFace safetensors omits frozen constants that the model constructor initializes\nRUN sed -i \"s/cfg.model.load_state_dict(state)/cfg.model.load_state_dict(state, strict=False)/\" tools/deployment/export_onnx.py\nRUN python3 tools/deployment/export_onnx.py -c configs/deimv2/deimv2_${BACKBONE}_${MODEL_SIZE}_coco.yml -r output/deimv2.pth\nFROM scratch\nARG BACKBONE\nARG MODEL_SIZE\nCOPY --from=build /deimv2/output/deimv2.onnx /deimv2_${BACKBONE}_${MODEL_SIZE}.onnx\nEOF\n```",
"ui": "Navigate to **Settings > System > Detectors and model** and select **ONNX** from the detector type dropdown and click **Add**. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | ------------------------------------------- |\n| **Object Detection Model Type** | `dfine` |\n| **Object detection model input width** | `640` |\n| **Object detection model input height** | `640` |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input D Type** | `float` |\n| **Custom object detector model path** | `/config/model_cache/dfine_m_obj2coco.onnx` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors:\n onnx:\n type: onnx\n\nmodel:\n model_type: dfine\n width: 640\n height: 640\n input_tensor: nchw\n input_dtype: float\n path: /config/model_cache/dfine_m_obj2coco.onnx\n labelmap_path: /labelmap/coco-80.txt"
},
{
"key": "yolo-legacy",
"label": "YOLO (v3, v4, v7)",
"recommended": false,
"download": "To export as ONNX:\n\n```sh\ngit clone https://github.com/NateMeyer/tensorrt_demos\ncd tensorrt_demos/yolo\n./download_yolo.sh\npython3 yolo_to_onnx.py -m yolov7-320\n```",
"ui": "Navigate to **Settings > System > Detectors and model** and select **ONNX** from the detector type dropdown and click **Add**. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | -------------------------------------------------------- |\n| **Object Detection Model Type** | `yolo-generic` |\n| **Object detection model input width** | `320` (should match the imgsize set during model export) |\n| **Object detection model input height** | `320` (should match the imgsize set during model export) |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input D Type** | `float` |\n| **Custom object detector model path** | `/config/model_cache/yolo.onnx` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors:\n onnx:\n type: onnx\n\nmodel:\n model_type: yolo-generic\n width: 320 # <--- should match the imgsize set during model export\n height: 320 # <--- should match the imgsize set during model export\n input_tensor: nchw\n input_dtype: float\n path: /config/model_cache/yolo.onnx\n labelmap_path: /labelmap/coco-80.txt"
}
]
},
"cpu": {
"title": "CPU",
"models": [
{
"key": "ssd",
"label": "MobileNet v2",
"recommended": true,
"download": "A TensorFlow Lite model is provided in the container at `/cpu_model.tflite` and is used by this detector type by default. To provide your own model, bind mount the file into the container and provide the path with `model.path`.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **CPU** from the detector type dropdown and click **Add**. Configure the number of threads and click **Add** again to add additional CPU detectors as needed (one per camera is recommended).\n\n| Field | Value |\n| ----------------- | ----- |\n| **Detector type** | `cpu` |\n| **Num threads** | `3` |",
"yaml": "detectors:\n cpu1:\n type: cpu\n num_threads: 3"
}
]
},
"deepstack": {
"title": "DeepStack / CodeProject.AI",
"models": [
{
"key": "yolo",
"label": "YOLO",
"recommended": true,
"download": "This detector runs object detection over the network against a CodeProject.AI or DeepStack server, so no model is downloaded into Frigate itself. Visit the [CodeProject.AI official website](https://www.codeproject.com/Articles/5322557/CodeProject-AI-Server-AI-the-easy-way) to download and install the AI server on your preferred device (e.g. Raspberry Pi, Nvidia Jetson, or other compatible hardware) before configuring the detector.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **DeepStack** from the detector type dropdown and click **Add**. Set the API URL to point to your CodeProject.AI server (e.g., `http://<your_codeproject_ai_server_ip>:<port>/v1/vision/detection`).\n\n| Field | Value |\n| ------------- | ---------------------------------------------------------------------- |\n| **API URL** | `http://<your_codeproject_ai_server_ip>:<port>/v1/vision/detection` |\n| **API Timeout** | `0.1` (seconds) |",
"yaml": "detectors:\n deepstack:\n api_url: http://<your_codeproject_ai_server_ip>:<port>/v1/vision/detection\n type: deepstack\n api_timeout: 0.1 # seconds"
}
]
},
"memryx": {
"title": "MemryX",
"models": [
{
"key": "yolonas",
"label": "YOLO-NAS",
"recommended": true,
"download": "The [YOLO-NAS](https://github.com/Deci-AI/super-gradients/blob/master/YOLONAS.md) model included in this detector is downloaded automatically and compiled to DFP with [mx_nc](https://developer.memryx.com/2p1/tools/neural_compiler.html#usage).\n\n**Note:** The default model for the MemryX detector is YOLO-NAS 320x320.\n\nThe input size for **YOLO-NAS** can be set to either **320x320** (default) or **640x640**.\n\n- The default size of **320x320** is optimized for lower CPU usage and faster inference times.\n\nMemryX `.dfp` models are automatically downloaded at runtime, if enabled, to the container at `/memryx_models/model_folder/`.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **MemryX** from the detector type dropdown and click **Add**, then set device to `PCIe:0`. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | ------------------------------------------------- |\n| **Object Detection Model Type** | `yolonas` |\n| **Object detection model input width** | `320` (can be set to `640` for higher resolution) |\n| **Object detection model input height** | `320` (can be set to `640` for higher resolution) |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input D Type** | `float` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors:\n memx0:\n type: memryx\n device: PCIe:0\n\nmodel:\n model_type: yolonas\n width: 320 # (Can be set to 640 for higher resolution)\n height: 320 # (Can be set to 640 for higher resolution)\n input_tensor: nchw\n input_dtype: float\n labelmap_path: /labelmap/coco-80.txt\n # Optional: The model is normally fetched through the runtime, so 'path' can be omitted unless you want to use a custom or local model.\n # path: /config/yolonas.zip\n # The .zip file must contain:\n # \u251c\u2500\u2500 yolonas.dfp (a file ending with .dfp)\n # \u2514\u2500\u2500 yolonas_post.onnx (optional; only if the model includes a cropped post-processing network)"
},
{
"key": "yolov9",
"label": "YOLOv9",
"recommended": false,
"download": "The YOLOv9s model included in this detector is downloaded from [the original GitHub](https://github.com/WongKinYiu/yolov9) and compiled to DFP with [mx_nc](https://developer.memryx.com/2p1/tools/neural_compiler.html#usage).\n\nMemryX `.dfp` models are automatically downloaded at runtime, if enabled, to the container at `/memryx_models/model_folder/`.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **MemryX** from the detector type dropdown and click **Add**, then set device to `PCIe:0`. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | ------------------------------------------------- |\n| **Object Detection Model Type** | `yolo-generic` |\n| **Object detection model input width** | `320` (can be set to `640` for higher resolution) |\n| **Object detection model input height** | `320` (can be set to `640` for higher resolution) |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input D Type** | `float` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors:\n memx0:\n type: memryx\n device: PCIe:0\n\nmodel:\n model_type: yolo-generic\n width: 320 # (Can be set to 640 for higher resolution)\n height: 320 # (Can be set to 640 for higher resolution)\n input_tensor: nchw\n input_dtype: float\n labelmap_path: /labelmap/coco-80.txt\n # Optional: The model is normally fetched through the runtime, so 'path' can be omitted unless you want to use a custom or local model.\n # path: /config/yolov9.zip\n # The .zip file must contain:\n # \u251c\u2500\u2500 yolov9.dfp (a file ending with .dfp)"
},
{
"key": "yolox",
"label": "YOLOX",
"recommended": false,
"download": "The model is sourced from the [OpenCV Model Zoo](https://github.com/opencv/opencv_zoo) and precompiled to DFP.\n\nMemryX `.dfp` models are automatically downloaded at runtime, if enabled, to the container at `/memryx_models/model_folder/`.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **MemryX** from the detector type dropdown and click **Add**, then set device to `PCIe:0`. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | ----------------------- |\n| **Object Detection Model Type** | `yolox` |\n| **Object detection model input width** | `640` |\n| **Object detection model input height** | `640` |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input D Type** | `float_denorm` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors:\n memx0:\n type: memryx\n device: PCIe:0\n\nmodel:\n model_type: yolox\n width: 640\n height: 640\n input_tensor: nchw\n input_dtype: float_denorm\n labelmap_path: /labelmap/coco-80.txt\n # Optional: The model is normally fetched through the runtime, so 'path' can be omitted unless you want to use a custom or local model.\n # path: /config/yolox.zip\n # The .zip file must contain:\n # \u251c\u2500\u2500 yolox.dfp (a file ending with .dfp)"
},
{
"key": "ssd",
"label": "SSDLite MobileNet v2",
"recommended": false,
"download": "The model is sourced from the [OpenMMLab Model Zoo](https://mmdeploy-oss.openmmlab.com/model/mmdet-det/ssdlite-e8679f.onnx) and has been converted to DFP.\n\nMemryX `.dfp` models are automatically downloaded at runtime, if enabled, to the container at `/memryx_models/model_folder/`.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **MemryX** from the detector type dropdown and click **Add**, then set device to `PCIe:0`. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | ----------------------- |\n| **Object Detection Model Type** | `ssd` |\n| **Object detection model input width** | `320` |\n| **Object detection model input height** | `320` |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input D Type** | `float` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors:\n memx0:\n type: memryx\n device: PCIe:0\n\nmodel:\n model_type: ssd\n width: 320\n height: 320\n input_tensor: nchw\n input_dtype: float\n labelmap_path: /labelmap/coco-80.txt\n # Optional: The model is normally fetched through the runtime, so 'path' can be omitted unless you want to use a custom or local model.\n # path: /config/ssdlite_mobilenet.zip\n # The .zip file must contain:\n # \u251c\u2500\u2500 ssdlite_mobilenet.dfp (a file ending with .dfp)\n # \u2514\u2500\u2500 ssdlite_mobilenet_post.onnx (optional; only if the model includes a cropped post-processing network)"
}
]
},
"tensorrt": {
"title": "TensorRT",
"models": [
{
"key": "yolo-legacy",
"label": "YOLO (v3, v4, v7)",
"recommended": true,
"download": "The model used for TensorRT must be preprocessed on the same hardware platform that it will run on, so Frigate generates the `.trt` model file on-device at startup. Processed models are stored in the `/config/model_cache` folder.\n\nBy default no models are generated. Set the `YOLO_MODELS` environment variable in Docker to one or more comma-separated model names (from the available `yolov3`/`yolov4`/`yolov7` models) and each one will be generated on startup if the corresponding `{model}.trt` file is not already present in `model_cache` (delete it to force regeneration). On Jetson devices with DLAs (Xavier or Orin), append `-dla` to a model name to generate a DLA model. If your GPU does not support FP16 operations, pass `USE_FP16=False` to disable it.\n\nAn example `docker-compose.yml` fragment that converts the `yolov7-320` and `yolov7x-640` models:\n\n```yml\nfrigate:\n environment:\n - YOLO_MODELS=yolov7-320,yolov7x-640\n - USE_FP16=false\n```",
"ui": "Navigate to **Settings > System > Detectors and model** and select **TensorRT** from the detector type dropdown and click **Add**, then set the device to `0` (the default GPU index). Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | ------------------------------------------------------------ |\n| **Custom object detector model path** | `/config/model_cache/tensorrt/yolov7-320.trt` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |\n| **Model Input Tensor Shape** | `nchw` |\n| **Model Input Pixel Color Format** | `rgb` |\n| **Object detection model input width** | `320` (MUST match the chosen model, e.g., yolov7-320 -> 320) |\n| **Object detection model input height** | `320` (MUST match the chosen model, e.g., yolov7-320 -> 320) |",
"yaml": "detectors:\n tensorrt:\n type: tensorrt\n device: 0 #This is the default, select the first GPU\n\nmodel:\n path: /config/model_cache/tensorrt/yolov7-320.trt\n labelmap_path: /labelmap/coco-80.txt\n input_tensor: nchw\n input_pixel_format: rgb\n width: 320 # MUST match the chosen model i.e yolov7-320 -> 320, yolov4-416 -> 416\n height: 320 # MUST match the chosen model i.e yolov7-320 -> 320 yolov4-416 -> 416"
}
]
},
"synaptics": {
"title": "Synaptics",
"models": [
{
"key": "ssd",
"label": "SSD MobileNet",
"recommended": true,
"download": "A synap model is provided in the container at `/mobilenet.synap` and is used by this detector type by default. The model comes from the [Synap-release Github](https://github.com/synaptics-astra/synap-release/tree/v1.5.0/models/dolphin/object_detection/coco/model/mobilenet224_full80).",
"ui": "Navigate to **Settings > System > Detectors and model** and select **Synaptics** from the detector type dropdown and click **Add**. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | ---------------------------- |\n| **Custom object detector model path** | `/synaptics/mobilenet.synap` |\n| **Object detection model input width** | `224` |\n| **Object detection model input height** | `224` |\n| **Model Input Tensor Shape** | `nhwc` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors: # required\n synap_npu: # required\n type: synaptics # required\n\nmodel: # required\n path: /synaptics/mobilenet.synap # required\n width: 224 # required\n height: 224 # required\n input_tensor: nhwc # default value (optional. If you change the model, it is required)\n labelmap_path: /labelmap/coco-80.txt # required"
}
]
},
"rknn": {
"title": "RKNN",
"models": [
{
"key": "yolov9",
"label": "YOLOv9",
"recommended": true,
"download": "If no custom model is provided, the RKNN detector downloads a default model from GitHub on first startup. Once cached, the model works fully offline. All models are automatically downloaded and stored in the folder `config/model_cache/rknn_cache`. After upgrading Frigate, you should remove older models to free up space.\n\nYou can also provide your own `.rknn` model. You should not save your own models in the `rknn_cache` folder, store them directly in the `model_cache` folder or another subfolder. To convert a model to `.rknn` format see the `rknn-toolkit2` (requires a x86 machine). Note, that there is only post-processing for the supported models.",
"ui": "Navigate to **Settings > System > Detectors and model** and, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | -------------------------------------------------- |\n| **Custom object detector model path** | `frigate-fp16-yolov9-t` (or other yolov9 variants) |\n| **Object Detection Model Type** | `yolo-generic` |\n| **Object detection model input width** | `320` |\n| **Object detection model input height** | `320` |\n| **Model Input Tensor Shape** | `nhwc` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "model: # required\n # name of model (will be automatically downloaded) or path to your own .rknn model file\n # possible values are:\n # - frigate-fp16-yolov9-t\n # - frigate-fp16-yolov9-s\n # - frigate-fp16-yolov9-m\n # - frigate-fp16-yolov9-c\n # - frigate-fp16-yolov9-e\n # your yolo_model.rknn\n path: frigate-fp16-yolov9-t\n model_type: yolo-generic\n width: 320\n height: 320\n input_tensor: nhwc\n labelmap_path: /labelmap/coco-80.txt"
},
{
"key": "yolonas",
"label": "YOLO-NAS",
"recommended": false,
"download": "If no custom model is provided, the RKNN detector downloads a default model from GitHub on first startup. Once cached, the model works fully offline. All models are automatically downloaded and stored in the folder `config/model_cache/rknn_cache`. After upgrading Frigate, you should remove older models to free up space.\n\nYou can also provide your own `.rknn` model. You should not save your own models in the `rknn_cache` folder, store them directly in the `model_cache` folder or another subfolder. To convert a model to `.rknn` format see the `rknn-toolkit2` (requires a x86 machine). Note, that there is only post-processing for the supported models.\n\n**Note:** The pre-trained YOLO-NAS weights from DeciAI are subject to their license and can't be used commercially. For more information, see: https://docs.deci.ai/super-gradients/latest/LICENSE.YOLONAS.html",
"ui": "Navigate to **Settings > System > Detectors and model** and, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | ----------------------------------------------------------------------- |\n| **Custom object detector model path** | `deci-fp16-yolonas_s` (or `deci-fp16-yolonas_m`, `deci-fp16-yolonas_l`) |\n| **Object Detection Model Type** | `yolonas` |\n| **Object detection model input width** | `320` |\n| **Object detection model input height** | `320` |\n| **Model Input Pixel Color Format** | `bgr` |\n| **Model Input Tensor Shape** | `nhwc` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "model: # required\n # name of model (will be automatically downloaded) or path to your own .rknn model file\n # possible values are:\n # - deci-fp16-yolonas_s\n # - deci-fp16-yolonas_m\n # - deci-fp16-yolonas_l\n # your yolonas_model.rknn\n path: deci-fp16-yolonas_s\n model_type: yolonas\n width: 320\n height: 320\n input_pixel_format: bgr\n input_tensor: nhwc\n labelmap_path: /labelmap/coco-80.txt"
},
{
"key": "yolox",
"label": "YOLOx",
"recommended": false,
"download": "If no custom model is provided, the RKNN detector downloads a default model from GitHub on first startup. Once cached, the model works fully offline. All models are automatically downloaded and stored in the folder `config/model_cache/rknn_cache`. After upgrading Frigate, you should remove older models to free up space.\n\nYou can also provide your own `.rknn` model. You should not save your own models in the `rknn_cache` folder, store them directly in the `model_cache` folder or another subfolder. To convert a model to `.rknn` format see the `rknn-toolkit2` (requires a x86 machine). Note, that there is only post-processing for the supported models.",
"ui": "Navigate to **Settings > System > Detectors and model** and, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | ---------------------------------------------- |\n| **Custom object detector model path** | `rock-i8-yolox_nano` (or other yolox variants) |\n| **Object Detection Model Type** | `yolox` |\n| **Object detection model input width** | `416` |\n| **Object detection model input height** | `416` |\n| **Model Input Tensor Shape** | `nhwc` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "model: # required\n # name of model (will be automatically downloaded) or path to your own .rknn model file\n # possible values are:\n # - rock-i8-yolox_nano\n # - rock-i8-yolox_tiny\n # - rock-fp16-yolox_nano\n # - rock-fp16-yolox_tiny\n # your yolox_model.rknn\n path: rock-i8-yolox_nano\n model_type: yolox\n width: 416\n height: 416\n input_tensor: nhwc\n labelmap_path: /labelmap/coco-80.txt"
}
]
},
"axengine": {
"title": "AXEngine",
"models": [
{
"key": "yolov9",
"label": "YOLOv9",
"recommended": true,
"download": "A yolov9 axmodel is provided in the container at `/axmodels` and is used by this detector type by default. The AXEngine detector downloads its default model from HuggingFace on first startup; once cached, the model works fully offline.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **AXEngine NPU** from the detector type dropdown and click **Add**. Then on the same page, in the **Custom Model** tab, configure:\n\n| Field | Value |\n| ---------------------------------------- | ----------------------- |\n| **Custom object detector model path** | `frigate-yolov9-tiny` |\n| **Object Detection Model Type** | `yolo-generic` |\n| **Object detection model input width** | `320` |\n| **Object detection model input height** | `320` |\n| **Model Input D Type** | `int` |\n| **Model Input Pixel Color Format** | `bgr` |\n| **Label map for custom object detector** | `/labelmap/coco-80.txt` |",
"yaml": "detectors:\n axengine:\n type: axengine\n\nmodel:\n path: frigate-yolov9-tiny\n model_type: yolo-generic\n width: 320\n height: 320\n input_dtype: int\n input_pixel_format: bgr\n labelmap_path: /labelmap/coco-80.txt"
}
]
},
"degirumAiServer": {
"title": "DeGirum AI Server",
"models": [
{
"key": "ai-server-inference",
"label": "AI Server Inference",
"recommended": true,
"download": "Launch a DeGirum AI server as a Docker container, then point the detector at it. Add this to your `docker-compose.yml`:\n\n```yaml\ndegirum_detector:\n container_name: degirum\n image: degirum/aiserver:latest\n privileged: true\n ports:\n - \"8778:8778\"\n```\n\nSet `location` to the server's service name, container name, or `host:port`.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **DeGirum** from the detector type dropdown and click **Add**.\n\n| Field | Value |\n| --- | --- |\n| **Location** | `degirum` |\n| **Zoo** | `degirum/public` |\n| **Token** | your AI Hub token (optional for the public zoo) |\n",
"yaml": "degirum_detector:\n type: degirum\n location: degirum\n zoo: degirum/public\n token: dg_example_token\n"
}
]
},
"degirumLocal": {
"title": "DeGirum Local",
"models": [
{
"key": "local-inference",
"label": "Local Inference",
"recommended": true,
"download": "Run hardware directly inside the Frigate container with `@local`, removing the AI server hop. The matching device runtime (e.g. the Hailo runtime) must be installed in the container; confirm it with `degirum sys-info`.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **DeGirum** from the detector type dropdown and click **Add**.\n\n| Field | Value |\n| --- | --- |\n| **Location** | `@local` |\n| **Zoo** | `degirum/public` |\n| **Token** | your AI Hub token (optional for the public zoo) |\n",
"yaml": "degirum_detector:\n type: degirum\n location: @local\n zoo: degirum/public\n token: dg_example_token\n"
}
]
},
"degirumCloud": {
"title": "DeGirum AI Hub Cloud",
"models": [
{
"key": "ai-hub-cloud-inference",
"label": "AI Hub Cloud Inference",
"recommended": true,
"download": "Run inferences on DeGirum's [AI Hub](https://hub.degirum.com) cloud with `@cloud`. Sign up, create an access token, and set it as `token`. Network latency may require lowering your detection fps.",
"ui": "Navigate to **Settings > System > Detectors and model** and select **DeGirum** from the detector type dropdown and click **Add**.\n\n| Field | Value |\n| --- | --- |\n| **Location** | `@cloud` |\n| **Zoo** | `degirum/public` |\n| **Token** | your AI Hub token (optional for the public zoo) |\n",
"yaml": "degirum_detector:\n type: degirum\n location: @cloud\n zoo: degirum/public\n token: dg_example_token\n"
}
]
}
}
Reference in New Issue Copy Permalink