OWLv2Options

Options & Configuration · AiDotNet.VisionLanguage.Grounding

Configuration options for OWLv2: self-training for scaling open-vocabulary detection.

Inheritance: ModelOptions → GenerativeVLMOptions → GroundingVLMOptions → OWLv2Options

For Beginners

These options configure the OWLv2 model. Default values follow the original paper settings.

Constructors

Constructor Summary
OWLv2Options() Initializes a new instance with default values.
OWLv2Options(OWLv2Options other) Initializes a new instance by copying from another instance.

Properties

Property Summary
bool EnableSelfTraining { get; set; } Gets or sets whether self-training augmentation is enabled.
int DetectionDim { get; set; } Gets or sets the detection-decoder feature width.
int NumClassEmbeddings { get; set; } Gets or sets the width of the class-embedding head of the reference configuration.
int NumFusionLayers { get; set; } Gets or sets the number of cross-modal fusion layers.
int NumTextLayers { get; set; } Gets or sets the number of layers of the CLIP text transformer that embeds the text queries (12).
int PatchSize { get; set; } Gets or sets the side of the square image patches the ViT backbone embeds (14 for the ViT-L/14 backbone; 16 for B/16).
int TextEmbeddingDim { get; set; } Gets or sets the text-encoder feature width.