EoMTOptions

Options & Configuration · AiDotNet.ComputerVision.Segmentation.Foundation

Configuration options for the EoMT (Encoder-only Mask Transformer) model.

Inheritance: ModelOptions → NeuralNetworkOptions → ModelHyperparameterOptions → PanopticSegmentationOptions → EoMTOptions

For Beginners

EoMT removes the pixel and transformer decoders used by Mask2Former, placing mask queries directly inside a plain ViT (DINOv2). This yields 4.4x faster inference. The values here are the ones the paper publishes, so you can use the model without setting anything.

Constructors

Constructor Summary
EoMTOptions() Initializes a new instance carrying EoMT’s published defaults.
EoMTOptions(EoMTOptions other) Initializes a new instance by copying from another instance.

Properties

Property Summary
EoMTModelSize ModelSize { get; set; } Gets or sets the ViT encoder size variant.

Methods

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