Skip to main content

Class: ObjectsOptions

Defined in: src/world/objects/ObjectsOptions.ts:7

Configuration options for the ObjectDetector.

Constructors​

Constructor​

new ObjectsOptions(options?): ObjectsOptions

Defined in: src/world/objects/ObjectsOptions.ts:67

Parameters​

options?​
backendConfig?​

{ activeBackend?: "gemini" | "mediapipe"; gemini?: { generationConfig?: {[key: string]: unknown; }; responseSchema?: { items?: { properties?: ... | ...; required?: ... | ...; type?: ... | ...; }; type?: string; }; systemInstruction?: string; }; mediapipe?: { modelAssetPath?: string; scoreThreshold?: number; wasmFilesUrl?: string; }; } = ...

Configuration for the detection backends.

backendConfig.activeBackend?​

"gemini" | "mediapipe" = ...

The active backend to use for detection.

backendConfig.gemini?​

{ generationConfig?: {[key: string]: unknown; }; responseSchema?: { items?: { properties?: ... | ...; required?: ... | ...; type?: ... | ...; }; type?: string; }; systemInstruction?: string; } = ...

backendConfig.gemini.generationConfig?​

{[key: string]: unknown; } = ...

Extra Gemini generation config merged into the per-call config (over the SDK defaults). Use to pin sampling parameters such as temperature: 0 for deterministic detections.

backendConfig.gemini.responseSchema?​

{ items?: { properties?: ... | ...; required?: ... | ...; type?: ... | ...; }; type?: string; } = ...

backendConfig.gemini.responseSchema.items?​

{ properties?: ... | ...; required?: ... | ...; type?: ... | ...; } = ...

backendConfig.gemini.responseSchema.items.properties?​

... | ... = ...

backendConfig.gemini.responseSchema.items.required?​

... | ... = ...

backendConfig.gemini.responseSchema.items.type?​

... | ... = 'OBJECT'

backendConfig.gemini.responseSchema.type?​

string = 'ARRAY'

backendConfig.gemini.systemInstruction?​

string = ...

backendConfig.mediapipe?​

{ modelAssetPath?: string; scoreThreshold?: number; wasmFilesUrl?: string; } = ...

Configuration for MediaPipe backend.

backendConfig.mediapipe.modelAssetPath?​

string = 'https://storage.googleapis.com/mediapipe-models/object_detector/efficientdet_lite2/int8/latest/efficientdet_lite2.tflite'

backendConfig.mediapipe.scoreThreshold?​

number = 0.5

backendConfig.mediapipe.wasmFilesUrl?​

string = 'https://cdn.jsdelivr.net/npm/@mediapipe/tasks-vision@0.10.34/wasm'

debugging?​

boolean = false

enable?​

() => ObjectsOptions

enabled?​

boolean = false

objectImageMargin?​

number = 0.05

Margin to add when cropping the object image, as a percentage of image size.

pollingIntervalMs?​

number = 0

Minimum delay in milliseconds between continuous object detection runs. A value of 0 runs again as soon as the previous detection finishes.

showDebugVisualizations?​

boolean = false

simulatorOverride?​

boolean = false

Use simulator ground truth instead of a camera detector on desktop.

Returns​

ObjectsOptions

Properties​

backendConfig​

backendConfig: object

Defined in: src/world/objects/ObjectsOptions.ts:30

Configuration for the detection backends.

activeBackend​

activeBackend: "gemini" | "mediapipe"

The active backend to use for detection.

gemini​

gemini: object

gemini.generationConfig​

generationConfig: Record<string, unknown>

Extra Gemini generation config merged into the per-call config (over the SDK defaults). Use to pin sampling parameters such as temperature: 0 for deterministic detections.

gemini.responseSchema​

responseSchema: object

gemini.responseSchema.items​

items: object

gemini.responseSchema.items.properties​

properties: object

gemini.responseSchema.items.properties.objectName​

objectName: object

gemini.responseSchema.items.properties.objectName.type​

type: string = 'STRING'

gemini.responseSchema.items.properties.xmax​

xmax: object

gemini.responseSchema.items.properties.xmax.type​

type: string = 'NUMBER'

gemini.responseSchema.items.properties.xmin​

xmin: object

gemini.responseSchema.items.properties.xmin.type​

type: string = 'NUMBER'

gemini.responseSchema.items.properties.ymax​

ymax: object

gemini.responseSchema.items.properties.ymax.type​

type: string = 'NUMBER'

gemini.responseSchema.items.properties.ymin​

ymin: object

gemini.responseSchema.items.properties.ymin.type​

type: string = 'NUMBER'

gemini.responseSchema.items.required​

required: string[]

gemini.responseSchema.items.type​

type: string = 'OBJECT'

gemini.responseSchema.type​

type: string = 'ARRAY'

gemini.systemInstruction​

systemInstruction: string

mediapipe​

mediapipe: object

Configuration for MediaPipe backend.

mediapipe.modelAssetPath​

modelAssetPath: string = 'https://storage.googleapis.com/mediapipe-models/object_detector/efficientdet_lite2/int8/latest/efficientdet_lite2.tflite'

mediapipe.scoreThreshold​

scoreThreshold: number = 0.5

mediapipe.wasmFilesUrl​

wasmFilesUrl: string = 'https://cdn.jsdelivr.net/npm/@mediapipe/tasks-vision@0.10.34/wasm'

debugging​

debugging: boolean = false

Defined in: src/world/objects/ObjectsOptions.ts:8

enabled​

enabled: boolean = false

Defined in: src/world/objects/ObjectsOptions.ts:9

objectImageMargin​

objectImageMargin: number = 0.05

Defined in: src/world/objects/ObjectsOptions.ts:25

Margin to add when cropping the object image, as a percentage of image size.

pollingIntervalMs​

pollingIntervalMs: number = 0

Defined in: src/world/objects/ObjectsOptions.ts:19

Minimum delay in milliseconds between continuous object detection runs. A value of 0 runs again as soon as the previous detection finishes.

showDebugVisualizations​

showDebugVisualizations: boolean = false

Defined in: src/world/objects/ObjectsOptions.ts:10

simulatorOverride​

simulatorOverride: boolean = false

Defined in: src/world/objects/ObjectsOptions.ts:13

Use simulator ground truth instead of a camera detector on desktop.

Methods​

enable()​

enable(): ObjectsOptions

Defined in: src/world/objects/ObjectsOptions.ts:76

Enables the object detector.

Returns​

ObjectsOptions