PanopticSegmentationBase<T>

Base Classes · AiDotNet.ComputerVision.Segmentation.Common

Abstract base class for panoptic segmentation models that unify semantic and instance segmentation.

Inheritance: NeuralNetworkBase<T> → SegmentationModelBase<T> → PanopticSegmentationBase<T>

Implements: IPanopticSegmentation<T>

Known derived types: CUPS<T>, EoMT<T>, KMaXDeepLab<T>, Mask2Former<T>, MaskDINO<T>, MixedQueryTransformer<T>, ODISESegmentation<T>, ODISE<T>, OMGSeg<T>, OneFormer<T>, U2Seg<T>, UNINEXT<T>, XDecoder<T>

For Beginners

Panoptic segmentation gives you the most complete picture of a scene by combining semantic segmentation (labeling regions like “road”, “sky”) with instance segmentation (distinguishing individual objects like “car #1”, “car #2”).

Models extending this base class: Mask2Former, kMaX-DeepLab, OneFormer, ODISE.

Properties

Property Summary
int NumStuffClasses { get; }
int NumThingClasses { get; }

Methods

Method Summary
PanopticSegmentationResult<T> SegmentPanoptic(Tensor<T> image)