IPAdapterFaceIDPlusModel<T>

Models & Types · AiDotNet.Diffusion.Control

IP-Adapter FaceID Plus model for face-identity-preserving generation.

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

For Beginners

Give this model a photo of someone’s face, and it will generate new images that look like the same person in different poses, styles, or settings. It preserves the person’s identity using face recognition technology.

How It Works

Combines IP-Adapter’s image prompting with face recognition embeddings (FaceID) for identity-preserving face generation. Supports both face swapping and face-consistent character creation.

Reference: Ye et al., “IP-Adapter: Text Compatible Image Prompt Adapter”, 2023

Example

// Create IP-Adapter FaceID Plus for identity-preserving generation
var options = new DiffusionModelOptions<float>
{
    LatentChannels = 4,
    DefaultInferenceSteps = 30
};
            
// Generate new images preserving a person's face identity
var faceEmbedding = Tensor<float>.CreateRandom(new[] { 1, 512 });
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 IPAdapterFaceIDPlusModel<float>(options: options))
    .Build(trainX, trainY);
var prediction = result.Predict(faceEmbedding);

Constructors

Constructor Summary
IPAdapterFaceIDPlusModel(NeuralNetworkArchitecture<T> architecture, DiffusionModelOptions<T> options, INoiseScheduler<T> scheduler, UNetNoisePredictor<T> baseUNet, StandardVAE<T> vae, IConditioningModule<T> conditioner, double faceIdScale, int? seed) Initializes a new IP-Adapter FaceID Plus model.

Properties

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

Methods

Method Summary
ModelMetadata<T> GetModelMetadata()