OutpaintingMaskGenerator<T>

Models & Types · AiDotNet.Diffusion.MaskUtilities

Generates masks for outpainting by marking regions outside the original image bounds.

Implements: IDataTransformer<T, Tensor<T>, Tensor<T>>

For Beginners

Outpainting extends an image beyond its borders. This utility creates the mask that tells the model which parts are the original image (don’t change) and which parts need to be generated (extend). The feathering option makes the transition smoother.

How It Works

Creates a mask where the original image region is unmasked (0) and extended regions are masked (1). Supports extending in any direction with optional feathering at the boundary for smooth blending.

When used as an IDataTransformer, the Transform method takes the original image tensor and generates an outpainting mask based on the padding configured in the constructor.

Constructors

Constructor Summary
OutpaintingMaskGenerator(int featherRadius, int padTop, int padRight, int padBottom, int padLeft) Initializes a new instance of the OutpaintingMaskGenerator class.

Properties

Property Summary
bool IsFitted { get; }
bool SupportsInverseTransform { get; }
int[] ColumnIndices { get; }

Methods

Method Summary
Tensor<T> FitTransform(Tensor<T> data)
Tensor<T> Generate(int canvasHeight, int canvasWidth, int imageTop, int imageLeft, int imageHeight, int imageWidth) Generates an outpainting mask for the given canvas and image bounds.
Tensor<T> InverseTransform(Tensor<T> data)
Tensor<T> Transform(Tensor<T> data) Generates an outpainting mask from the input image tensor’s dimensions.
string[] GetFeatureNamesOut(string[] inputFeatureNames)
void Fit(Tensor<T> data)