nrtk.impls.score_detections.random_scorer
This module defines the RandomScorer class, which implements the ScoreDetections interface to generate random scores for object detection tasks. The RandomScorer class serves as a test tool to verify reproducibility by returning random float scores based on a given seed value. This can be useful in testing scenarios where deterministic random values are needed.
- Classes:
RandomScorer: An implementation of ScoreDetections that generates random scores for input detection data.
- Dependencies:
numpy for generating random float values.
smqtk_image_io for bounding box handling.
nrtk.interfaces for the ScoreDetections interface.
- Usage:
Instantiate RandomScorer with an optional seed for reproducibility:
scorer = RandomScorer(seed=42) scores = scorer.score(actual_detections, predicted_detections)
- Example:
random_scorer = RandomScorer(seed=123) random_scores = random_scorer.score(actual_detections, predicted_detections)
Classes
An implementation of the |