UniVSOptions
Options & Configuration · AiDotNet.ComputerVision.Segmentation.Video
Configuration options for UniVS universal video segmentation.
Inheritance: ModelOptions → NeuralNetworkOptions → ModelHyperparameterOptions → SegmentationModelOptions → UniVSOptions
For Beginners
These options configure the UniVS model. Default values follow the original paper settings.
Constructors
| Constructor | Summary |
|---|---|
UniVSOptions() |
Initializes a new instance with default values. |
UniVSOptions(UniVSOptions other) |
Initializes a new instance by copying from another instance. |
Properties
| Property | Summary |
|---|---|
UniVSModelSize ModelSize { get; set; } |
Gets or sets which published size variant of the model to build. |
int NumPromptPoints { get; set; } |
Gets or sets how many points the visual sampler takes from each visual prompt (a mask) to make its prompt tokens. |
int PromptMemoryLength { get; set; } |
Gets or sets how many of the most recent visual prompts each object’s prompt memory pool keeps. |
Methods
| Method | Summary |
|---|---|
void Validate() |
Throws when a value on this instance cannot produce a working model. |