Expand description
Semantic entities, embeddings, fusion, and nearest-neighbor search.
Structs§
- Embedding
- Dense float embedding.
- Entity
Id - Stable entity identifier.
- Fusion
Score - Weighted fusion of embedding similarity and label confidence.
- Multimodal
Fusion - Multimodal fusion weights.
- Onnx
Entity Embedder - Runs entity features through an already-open model session to produce an
Embedding. - Open
Vocab Label - Open-vocabulary label with confidence.
- Semantic
Entity - One semantic spatial entity with optional embedding.
- Semantic
Search Index - In-memory nearest-neighbor index over semantic entities.
- Spatial
Record Entity - A semantic entity derived from one versioned spatial record.
Enums§
- Semantic
Error - Failures in embeddings and semantic search.
Functions§
- cosine_
similarity - Cosine similarity in
[-1, 1]. - record_
entity_ id - Derives a stable entity id from a record’s protocol-independent lineage.
Type Aliases§
- Semantic
Result - Result type for semantic operations.