StyleAlignedEditModel<T>

Models & Types · AiDotNet.Diffusion.StyleTransfer

StyleAligned-Edit model for consistent multi-image style editing via shared self-attention.

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

For Beginners

When you need to edit multiple images with the same visual style (like a series of product photos or a consistent set of illustrations), StyleAligned-Edit ensures they all look like they belong together by sharing style information across images.

How It Works

StyleAligned-Edit extends the StyleAligned shared attention mechanism to image editing, ensuring that multiple edited images maintain visual consistency. All images in a batch share self-attention features, enforcing a coherent visual style across the set.

Reference: Hertz et al., “Style Aligned Image Generation via Shared Attention”, CVPR 2024

Example

var options = new DiffusionModelOptions<float> { LatentChannels = 4, DefaultInferenceSteps = 30 };
var input = Tensor<float>.CreateRandom(new[] { 1, 4, 64, 64 });
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 StyleAlignedEditModel<float>(options: options))
    .Build(trainX, trainY);
var edited = result.Predict(input);

Constructors

Constructor Summary
StyleAlignedEditModel(NeuralNetworkArchitecture<T> architecture, DiffusionModelOptions<T> options, INoiseScheduler<T> scheduler, UNetNoisePredictor<T> predictor, StandardVAE<T> vae, IConditioningModule<T> conditioner, 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()