EfficientNet<T>

Models & Types · AiDotNet.ComputerVision.Detection.Backbones

EfficientNet backbone for efficient feature extraction.

Inheritance: NeuralNetworkBase<T> → EfficientNet<T>

Implements: IDetectionBackbone<T>

How It Works

Reference: Tan et al., “EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks”, ICML 2019

Constructors

Constructor Summary
EfficientNet(EfficientNetBackboneOptions options) Creates a new EfficientNet backbone: the classifier’s trunk without its head convolution, producing the C2–C5 pyramid from the ends of stages 2, 3, 5 and 7 (strides 4, 8, 16 and 32), as EfficientDet reads it.

Properties

Property Summary
IReadOnlyList<int> OutputChannels { get; }
IReadOnlyList<int> Strides { get; }
bool IsFrozen { get; set; }
string Name { get; }

Methods

Method Summary
IFullModel<T, Tensor<T>, Tensor<T>> WithParameters(Vector<T> parameters)
IReadOnlyList<Tensor<T>> GetFeatureMaps(Tensor<T> input)
List<Tensor<T>> ExtractFeatures(Tensor<T> input)
ModelMetadata<T> GetModelMetadata()
ValueTuple<int, int> GetExpectedInputSize()
void Freeze()
void ReadParameters(BinaryReader reader)
void Unfreeze()
void WriteParameters(BinaryWriter writer)