GroundedSegment<T>

Models & Types · AiDotNet.VisionLanguage.Grounding

One region a caption names, with its SAM 2 mask.

Implements: IEquatable<GroundedSegment<T>>

Constructors

Constructor Summary
GroundedSegment(GroundingDetection<T> Detection, Tensor<T> Mask, T MaskScore) One region a caption names, with its SAM 2 mask.

Properties

Property Summary
GroundingDetection<T> Detection { get; set; } Grounding DINO’s box, score and phrase.
T MaskScore { get; set; } SAM 2’s predicted IoU for the mask.
Tensor<T> Mask { get; set; } The mask [H, W] at the image’s size: 1 inside, 0 outside.

Methods

Method Summary
GroundedSegment<T> <Clone>$()
bool Equals(GroundedSegment<T> other)
bool Equals(object obj)
bool op_Equality(GroundedSegment<T> left, GroundedSegment<T> right)
bool op_Inequality(GroundedSegment<T> left, GroundedSegment<T> right)
int GetHashCode()
string ToString()
void Deconstruct(GroundingDetection<T> Detection, Tensor<T> Mask, T MaskScore)