DADA<T>
Models & Types · AiDotNet.Augmentation.Image
DADA - Differentiable Automatic Data Augmentation (Li et al., 2020). Applies augmentations with learned probabilities via Gumbel-Softmax relaxation.
Inheritance: AugmentationBase<T, ImageTensor<T>> → ImageAugmenterBase<T> → DADA<T>
Constructors
| Constructor | Summary |
|---|---|
DADA(double temperature, int operationCount, double probability) |
Properties
| Property | Summary |
|---|---|
double Temperature { get; } |
|
int OperationCount { get; } |
Methods
| Method | Summary |
|---|---|
IDictionary<string, object> GetParameters() |