brails.processors.nFloor_vlm.nFloor_vlm module

Class object to predict number of floors from street imagery using CLIP.

class brails.processors.nFloor_vlm.nFloor_vlm.NFloorVLM(input_dict: dict | None = None)

Bases: CLIPClassifier

NFloorVLM classifier predicts number of floors in a building using CLIP.

This class is designed to initialize the classifier with text prompts and classes related to the number of floors and make predictions from street-level imagery.

Attributes__

text_prompts (List[str]): A list of text prompts for floor

classifications.

classes (List[int]): A list of classes representing the number of

floors (default [1, 2, 3])

Args__

input_dict (Optional[dict]): A dictionary containing prompts and

classes for customization.