Detection

Detection encapsulates object detection, model switching, and detection result drawing capabilities.

Signature:

Detection(
    model: str = "default",
    backend: str = "local",
    **kwargs: Any,
) -> Detection

Description:

Create and initialize a detector.

Parameters:

Parameter	Type	Default	Description
model	str	"default"	Model ID or local model path.
backend	str	"local"	Inference backend; currently supports "local".
device	str | None	None	Inference device, e.g., "cpu", "cuda:0".
engine	str | None	Model default	Local inference engine; supports "yolo", "onnx", "tensorrt".
providers	list[str] | None	None	ONNX Runtime provider list.

Returns:

Detection

Initialized detector instance.

Raises:

ValueError: Unsupported backend, engine, or model file.

ImportError: Missing dependencies for the selected inference engine.

RuntimeError: Failed to load inference engine model.

Example:

from boosteros.brain import Detection

detector = Detection(model="default", backend="local")