Enum OpenVocabularyModel
- Namespace
- VisioForge.Core.Types.X.AI
- Assembly
- VisioForge.Core.dll
The open-vocabulary detection model family a supplied ONNX model belongs to. This selects the text tokenizer, the frame-preprocessing convention, and the output decoder used by the open-vocabulary detector engine.
public enum OpenVocabularyModelFields
OWLv2 = 0-
OWLv2 (open-world localization). Uses the CLIP text tokenizer, pads the frame to a square (top-left anchored) and resizes it to 960x960, and decodes per-query sigmoid logits over a fixed 3600-patch grid.
GroundingDINO = 1-
Grounding DINO. Uses the BERT WordPiece tokenizer over a "prompt1 . prompt2 ." caption, resizes the frame to the model's fixed input size, and decodes per-query token-span scores over 900 queries.