DepthAnythingV2Options

Options & Configuration · AiDotNet.Video.Options

Configuration options for the Depth Anything V2 monocular depth model.

Inheritance: ModelOptions → NeuralNetworkOptions → ModelHyperparameterOptions → VideoHyperparameterOptions → DepthAnythingV2Options

For Beginners

This class carries the settings Depth Anything V2 was published with, so you can use the model without configuring anything. Pick a size if you want a different speed/accuracy trade-off:

How It Works

Choosing a size sets the feature width and encoder depth the paper pairs with it. You can still override either afterwards; the size is recorded but no longer consulted once set.

Constructors

Constructor Summary
DepthAnythingV2Options() Initializes a new instance carrying the Depth Anything V2 Base defaults.
DepthAnythingV2Options(DepthAnythingV2ModelSize modelSize) Initializes a new instance by copying another options instance.
DepthAnythingV2Options(DepthAnythingV2Options other) Initializes a new instance by copying another options instance.

Properties

Property Summary
DepthAnythingV2ModelSize ModelSize { get; set; } Gets or sets the encoder size variant.
int NumEncoderBlocks { get; set; } Gets or sets the number of DINOv2 transformer encoder blocks.
int PatchSize { get; set; } Gets or sets the side length in pixels of each image patch fed to the encoder.

Methods

Method Summary
void Validate() Throws if a value this model requires has been left unset or is not positive.