robokudo.annotators.clip_annotator¶
Classes¶
A simple CLIP annotator that uses a given vocabulary to annotate all incoming object hypothesis. |
Module Contents¶
- class robokudo.annotators.clip_annotator.ClipAnnotator(name: str = 'ClipAnnotator')¶
-
Bases:
robokudo.annotators.core.BaseAnnotatorA simple CLIP annotator that uses a given vocabulary to annotate all incoming object hypothesis.
- device¶
-
Torch device to be used by the annotator.
- clip_model¶
-
CLIP model used for feature extraction.
- clip_preprocess¶
-
CLIP preprocessor used for feature extraction.
- vocabulary: typing_extensions.List[str] = ['Milk Container', 'Conflakes', 'Pancake Batter', 'Instant Coffee', 'Crepe Pan', 'Dish Soap']¶
-
The vocabulary used for object classification
- vocabulary_index: faiss.IndexFlatIP¶
-
A faiss index for text features used in similarity search.
- update() py_trees.common.Status¶