HyperNeRFMetaAlgorithm<T, TInput, TOutput>
Models & Types · AiDotNet.MetaLearning.Algorithms
Implementation of HyperNeRF Meta: Positional-Encoding-Conditioned Meta-Learning.
Inheritance: ModelBase<T, TInput, TOutput> → MetaLearnerBase<T, TInput, TOutput> → HyperNeRFMetaAlgorithm<T, TInput, TOutput>
Implements: IGeneratedParameterRegistrar<T>
How It Works
Combines hypernetwork conditioning with NeRF-style sinusoidal positional encoding. Each parameter index is encoded using multi-frequency sin/cos functions, providing structural awareness of where parameters are located in the network. These positional features are combined with a task-specific latent code (from gradient statistics) to produce per-parameter learning rate modulation through a learned conditioning MLP.
Algorithm:
Constructors
| Constructor | Summary |
|---|---|
HyperNeRFMetaAlgorithm(HyperNeRFMetaOptions<T, TInput, TOutput> options) |
Properties
| Property | Summary |
|---|---|
MetaLearningAlgorithmType AlgorithmType { get; } |
Methods
| Method | Summary |
|---|---|
IModel<TInput, TOutput, ModelMetadata<T>> Adapt(IMetaLearningTask<T, TInput, TOutput> task) |
|
T MetaTrain(TaskBatch<T, TInput, TOutput> taskBatch) |