BridgeTowerOptions
Options & Configuration · AiDotNet.VisionLanguage.Foundational
Configuration options for BridgeTower cross-modal alignment model.
Inheritance: ModelOptions → FoundationalVLMOptions → BridgeTowerOptions
How It Works
BridgeTower (Xu et al., AAAI 2023) introduces bridge layers that connect vision and text encoder layers at multiple levels, enabling fine-grained cross-modal alignment. Each bridge layer consists of cross-attention between corresponding encoder layers.
Constructors
| Constructor | Summary |
|---|---|
BridgeTowerOptions() |
Initializes a new instance with default values. |
BridgeTowerOptions(BridgeTowerOptions other) |
Initializes a new instance by copying from another instance. |
Properties
| Property | Summary |
|---|---|
bool UseBidirectionalBridges { get; set; } |
Gets or sets whether to use bi-directional bridges. |
int BridgeDim { get; set; } |
Gets or sets the bridge layer hidden dimension. |
int NumBridgeLayers { get; set; } |
Gets or sets the number of bridge connection layers. |
int PatchSize { get; set; } |
Gets or sets the square visual patch size used by the ViT stem. |