DynamiCrafterOptions

Options & Configuration · AiDotNet.Video.Options

Configuration options for the DynamiCrafter video diffusion interpolation model.

Inheritance: ModelOptions → NeuralNetworkOptions → DynamiCrafterOptions

For Beginners

DynamiCrafter uses an AI image/video generator (diffusion model) that already knows how things move in the real world. Given a start frame and end frame, it gradually “imagines” what happens in between, producing natural-looking intermediate frames with realistic motion, lighting changes, and object interactions.

How It Works

DynamiCrafter (2024) uses video diffusion priors for frame interpolation:

  • Video diffusion backbone: adapts a pre-trained text-to-video diffusion model (e.g.,

Stable Video Diffusion) for the interpolation task, leveraging its learned motion priors

  • First/last frame conditioning: the diffusion process is conditioned on both the first

and last frames using CLIP image embeddings injected via cross-attention, ensuring the generated intermediate frames are temporally consistent with both endpoints

  • Noise schedule adaptation: modified diffusion noise schedule that biases early denoising

steps toward global motion consistency and later steps toward local detail refinement

  • Temporal attention: 3D self-attention across generated frames ensures smooth motion

transitions without flickering or temporal discontinuities

Constructors

Constructor Summary
DynamiCrafterOptions() Initializes a new instance with default values.
DynamiCrafterOptions(DynamiCrafterOptions other) Initializes a new instance by copying from another instance.

Properties

Property Summary
OnnxModelOptions OnnxOptions { get; set; } Gets or sets the ONNX runtime options.
VideoModelVariant Variant { get; set; } Gets or sets the model variant.
double DropoutRate { get; set; } Gets or sets the dropout rate.
double GuidanceScale { get; set; } Gets or sets the classifier-free guidance scale.
double LearningRate { get; set; } Gets or sets the learning rate.
int NumDiffusionSteps { get; set; } Gets or sets the number of diffusion timesteps during inference.
int NumFeatures { get; set; } Gets or sets the number of feature channels in the UNet.
int NumHeads { get; set; } Gets or sets the number of attention heads in temporal attention.
int NumResBlocks { get; set; } Gets or sets the number of UNet residual blocks per level.
string ModelPath { get; set; } Gets or sets the path to the ONNX model file.