ControlNetInpaintingModel<T>

Models & Types · AiDotNet.Diffusion.Control

ControlNet Inpainting model with mask-aware conditioning.

Inheritance: DiffusionModelBase<T> → LatentDiffusionModelBase<T> → ControlNetInpaintingModel<T>

For Beginners

This model erases part of an image (marked by a mask) and fills it in with new content that matches a control signal. For example, you could erase a person from a photo and use an edge map to guide what replaces them.

How It Works

Specialized ControlNet for inpainting that takes both a control image and a binary mask as input. The mask indicates which regions to regenerate while the control signal guides the structure of the inpainted content.

Example

// Create a ControlNet inpainting model
var options = new DiffusionModelOptions<float>
{
    LatentChannels = 4,
    DefaultInferenceSteps = 30
};
            
// Inpaint masked regions guided by a control image
var imageWithMask = Tensor<float>.CreateRandom(new[] { 1, 5, 64, 64 }); // 4 latent + 1 mask
var trainX = Tensor<float>.CreateRandom(4, 8);
var trainY = Tensor<float>.CreateRandom(4, 2);
var result = new AiModelBuilder<float, Tensor<float>, Tensor<float>>()
    .ConfigureModel(new ControlNetInpaintingModel<float>(options: options, controlType: ControlType.Canny))
    .Build(trainX, trainY);
var prediction = result.Predict(imageWithMask);

Constructors

Constructor Summary
ControlNetInpaintingModel(NeuralNetworkArchitecture<T> architecture, DiffusionModelOptions<T> options, INoiseScheduler<T> scheduler, UNetNoisePredictor<T> baseUNet, StandardVAE<T> vae, IConditioningModule<T> conditioner, ControlType controlType, int? seed)

Properties

Property Summary
IConditioningModule<T> Conditioner { get; }
INoisePredictor<T> NoisePredictor { get; }
IVAEModel<T> VAE { get; }
int LatentChannels { get; }

Methods

Method Summary
ModelMetadata<T> GetModelMetadata()