Ensure that your bounding boxes comply with the bounding box spec

• Original Link : https://keras.io/api/keras_cv/bounding_box/utils/validate_format/
• Last Checked at : 2024-11-25

source

validate_format function

keras_cv.bounding_box.validate_format(bounding_boxes, variable_name="bounding_boxes")

validates that a given set of bounding boxes complies with KerasCV format.

For a set of bounding boxes to be valid it must satisfy the following conditions:

• bounding_boxes must be a dictionary
• contains keys "boxes" and "classes"
• each entry must have matching first two dimensions; representing the batch axis and the number of boxes per image axis.
• either both "boxes" and "classes" are batched, or both are unbatched.

Additionally, one of the following must be satisfied:

• "boxes" and "classes" are both Ragged
• "boxes" and "classes" are both Dense
• "boxes" and "classes" are unbatched

Arguments

• bounding_boxes: dictionary of bounding boxes according to KerasCV format.

Raises

ValueError if any of the above conditions are not met