Interface TensorFactory

type TensorFactory defines the factory functions of 'Tensor' to create tensor instances from existing data or resources.

interface TensorFactory {
    fromGpuBuffer<T>(buffer, options): TypedTensor<T>;
    fromImage(imageData, options?): Promise<TypedTensor<"float32"> | TypedTensor<"uint8">>;
    fromImage(imageElement, options?): Promise<TypedTensor<"float32"> | TypedTensor<"uint8">>;
    fromImage(urlSource, options?): Promise<TypedTensor<"float32"> | TypedTensor<"uint8">>;
    fromImage(bitmap, options): Promise<TypedTensor<"float32"> | TypedTensor<"uint8">>;
    fromMLTensor<T>(tensor, options): TypedTensor<T>;
    fromPinnedBuffer<T>(type, buffer, dims?): TypedTensor<T>;
    fromTexture<T>(texture, options): TypedTensor<"float32">;
}

Hierarchy (view full)

Methods

fromGpuBuffer fromImage fromMLTensor fromPinnedBuffer fromTexture

Methods

fromGpuBuffer

  • fromGpuBuffer<T>(buffer, options): TypedTensor<T>

  • create a tensor from a WebGPU buffer

    Type Parameters

    Parameters

    • buffer: GpuBufferTypeFallback

      the GPUBuffer object to create tensor from

    • options: TensorFromGpuBufferOptions<T>

      An optional object representing options for creating tensor from WebGPU buffer.

      The options include following properties:

      • dataType: the data type of the tensor. If omitted, assume 'float32'.
      • dims: the dimension of the tensor. Required.
      • download: an optional function to download the tensor data from GPU to CPU. If omitted, the GPU data will not be able to download. Usually, this is provided by a GPU backend for the inference outputs. Users don't need to provide this function.
      • dispose: an optional function to dispose the tensor data on GPU. If omitted, the GPU data will not be disposed. Usually, this is provided by a GPU backend for the inference outputs. Users don't need to provide this function.

    Returns TypedTensor<T>

    a tensor object

fromImage

  • fromImage(imageData, options?): Promise<TypedTensor<"float32"> | TypedTensor<"uint8">>

  • create a tensor from an ImageData object

    Parameters

    • imageData: ImageData

      the ImageData object to create tensor from

    • Optional options: TensorFromImageDataOptions

      An optional object representing options for creating tensor from ImageData.

      The following default settings will be applied:

      • tensorFormat: 'RGB'
      • tensorLayout: 'NCHW'
      • dataType: 'float32'

    Returns Promise<TypedTensor<"float32"> | TypedTensor<"uint8">>

    A promise that resolves to a tensor object

  • fromImage(imageElement, options?): Promise<TypedTensor<"float32"> | TypedTensor<"uint8">>

  • create a tensor from a HTMLImageElement object

    Parameters

    • imageElement: HTMLImageElement

      the HTMLImageElement object to create tensor from

    • Optional options: TensorFromImageElementOptions

      An optional object representing options for creating tensor from HTMLImageElement.

      The following default settings will be applied:

      • tensorFormat: 'RGB'
      • tensorLayout: 'NCHW'
      • dataType: 'float32'

    Returns Promise<TypedTensor<"float32"> | TypedTensor<"uint8">>

    A promise that resolves to a tensor object

  • fromImage(urlSource, options?): Promise<TypedTensor<"float32"> | TypedTensor<"uint8">>

  • create a tensor from URL

    Parameters

    • urlSource: string

      a string as a URL to the image or a data URL containing the image data.

    • Optional options: TensorFromUrlOptions

      An optional object representing options for creating tensor from URL.

      The following default settings will be applied:

      • tensorFormat: 'RGB'
      • tensorLayout: 'NCHW'
      • dataType: 'float32'

    Returns Promise<TypedTensor<"float32"> | TypedTensor<"uint8">>

    A promise that resolves to a tensor object

  • fromImage(bitmap, options): Promise<TypedTensor<"float32"> | TypedTensor<"uint8">>

  • create a tensor from an ImageBitmap object

    Parameters

    • bitmap: ImageBitmap

      the ImageBitmap object to create tensor from

    • options: TensorFromImageBitmapOptions

      An optional object representing options for creating tensor from URL.

      The following default settings will be applied:

      • tensorFormat: 'RGB'
      • tensorLayout: 'NCHW'
      • dataType: 'float32'

    Returns Promise<TypedTensor<"float32"> | TypedTensor<"uint8">>

    A promise that resolves to a tensor object

fromMLTensor

  • fromMLTensor<T>(tensor, options): TypedTensor<T>

  • create a tensor from a WebNN MLTensor

    Type Parameters

    Parameters

    • tensor: MLTensorTypeFallback

      the MLTensor object to create tensor from

    • options: TensorFromMLTensorOptions<T>

      An optional object representing options for creating tensor from a WebNN MLTensor.

      The options include following properties:

      • dataType: the data type of the tensor. If omitted, assume 'float32'.
      • dims: the dimension of the tensor. Required.
      • download: an optional function to download the tensor data from the MLTensor to CPU. If omitted, the MLTensor data will not be able to download. Usually, this is provided by the WebNN backend for the inference outputs. Users don't need to provide this function.
      • dispose: an optional function to dispose the tensor data on the WebNN MLTensor. If omitted, the MLTensor will not be disposed. Usually, this is provided by the WebNN backend for the inference outputs. Users don't need to provide this function.

    Returns TypedTensor<T>

    a tensor object

fromPinnedBuffer

  • fromPinnedBuffer<T>(type, buffer, dims?): TypedTensor<T>

  • create a tensor from a pre-allocated buffer. The buffer will be used as a pinned buffer.

    Type Parameters

    • T extends "float32" | "uint8" | "int8" | "uint16" | "int16" | "int32" | "int64" | "bool" | "float16" | "float64" | "uint32" | "uint64" | "uint4" | "int4"

    Parameters

    • type: T

      the tensor element type.

    • buffer: DataTypeMap[T]

      a TypedArray corresponding to the type.

    • Optional dims: readonly number[]

      specify the dimension of the tensor. If omitted, a 1-D tensor is assumed.

    Returns TypedTensor<T>

    a tensor object

fromTexture

  • fromTexture<T>(texture, options): TypedTensor<"float32">

  • create a tensor from a WebGL texture

    Type Parameters

    • T extends "float32" = "float32"

    Parameters

    • texture: WebGLTexture

      the WebGLTexture object to create tensor from

    • options: TensorFromTextureOptions<T>

      An optional object representing options for creating tensor from WebGL texture.

      The options include following properties:

      • width: the width of the texture. Required.
      • height: the height of the texture. Required.
      • format: the format of the texture. If omitted, assume 'RGBA'.
      • download: an optional function to download the tensor data from GPU to CPU. If omitted, the GPU data will not be able to download. Usually, this is provided by a GPU backend for the inference outputs. Users don't need to provide this function.
      • dispose: an optional function to dispose the tensor data on GPU. If omitted, the GPU data will not be disposed. Usually, this is provided by a GPU backend for the inference outputs. Users don't need to provide this function.

    Returns TypedTensor<"float32">

    a tensor object

Settings

Member Visibility

Theme

On This Page

Generated using TypeDoc