Detector

Runs on full frames and produces bounding box detections. At least one primary detector (without operate_on) is required in every pipeline.

models:
  - name: "yolo-detector"
    type: "detector"
    architecture: "yolo11"
    onnx_file: /path/to/model.onnx
    label_file: /path/to/labels.txt

Fields

Field	Type	Required	Default	Description
architecture	string	Yes	—	Model architecture (see supported architectures)
active_classes	list of int	No	all classes	Only detect these class IDs (0-indexed from label file)
operate_on	string	No	—	Name of parent model (makes this a secondary detector)
operate_on_classes	list of int	No	all parent classes	Only process objects of these class IDs from the parent

Plus the common model fields (name, onnx_file, label_file, batch_size, precision).

Supported architectures

Architecture	Description
yolo5	YOLOv5
yolo6	YOLOv6
yolo7	YOLOv7
yolo8	YOLOv8
yolo9	YOLOv9
yolo10	YOLOv10 (built-in NMS)
yolo11	YOLO11
yolo12	YOLOv12
yolo13	YOLOv13
yolo26	YOLO26 (built-in NMS)
yolox	YOLOX
detectnet_v2	NVIDIA DetectNet_v2 (TAO / ResNet-based)