Skip to content

ObjectDetectorModel

Defined in: packages/runntime/src/zoo/tasks/cv/objectDetection.ts:24

What to load. Every field is optional; a missing modelPath comes from models.objectDetection.YOLO26.DEFAULT.

readonly optional inputSize?: number

Defined in: packages/runntime/src/zoo/tasks/cv/objectDetection.ts:31

Input width and height in pixels, a multiple of 32. Default 640. Smaller runs faster and misses more small objects.


readonly optional modelPath?: ModelPath

Defined in: packages/runntime/src/zoo/tasks/cv/objectDetection.ts:26

The weights: a safetensors export of a YOLO26 detect checkpoint.


readonly optional resizeMode?: ResizeMode

Defined in: packages/runntime/src/zoo/tasks/cv/objectDetection.ts:33

How an image is fitted into the model input. Default letterbox.


readonly optional variant?: "x" | "n" | "m" | "s" | "l"

Defined in: packages/runntime/src/zoo/tasks/cv/objectDetection.ts:28

Model size, n to x. Default: read from the weights.