def _eval(self, db_segmentation, eval_func, measure, scale=1):
annotations = self._masks[1:-1]
# Strip of first and last frame if available
segmentation = db_segmentation._masks[1 - db_segmentation.
_frames[0]:len(annotations) + 1 -
db_segmentation._frames[0]]
assert len(annotations) == len(segmentation)
if measure == 'T':
magic_number = 5.0
X = np.array([np.nan] + [
eval_func(an, sg) * magic_number
for an, sg in zip(segmentation[:-1], segmentation[1:])
] + [np.nan])
else:
X = np.array([np.nan] + [
eval_func(an, sg) for an, sg in zip(annotations, segmentation)
] + [np.nan])
from utils import db_statistics
M, O, D = db_statistics(X)
if measure == 'T': O = D = np.nan
return X, M, O, D
def _eval(self,db_segmentation,eval_func,measure,scale=1): annotations = self._masks[1:-1] # Strip of first and last frame if available segmentation = db_segmentation._masks[ 1-db_segmentation._frames[0]:len(annotations)+1-db_segmentation._frames[0]] assert len(annotations) == len(segmentation) if measure == 'T': magic_number = 5.0 X = np.array([np.nan]+[eval_func(an,sg)*magic_number for an,sg in zip(segmentation[:-1],segmentation[1:])] + [np.nan]) else: X = np.array([np.nan]+[eval_func(an,sg) for an,sg in zip(annotations,segmentation)] + [np.nan]) from utils import db_statistics M,O,D = db_statistics(X) if measure == 'T': O = D = np.nan return X,M,O,D