nrtk.impls.gen_object_detector_blackbox_response.simple_generic_generator
Defines SimpleGenericGenerator to generate item-response curves from images and ground-truth boxes.
- Classes:
SimpleGenericGenerator: An example implementation that generates object detection responses for a sequence of images and their associated ground-truth data.
- Dependencies:
numpy for handling image data.
smqtk_image_io for bounding box operations.
nrtk.interfaces for blackbox response generation interface.
- Example usage:
images = [image1, image2, …] ground_truth = [
[(bbox1, {label1: score1}), (bbox2, {label2: score2})], …
] generator = SimpleGenericGenerator(images, ground_truth) image, ground_truth, metadata = generator[0]
Classes
Example implementation of the |