Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Adding Gaussian Blur as 'hide_color' option #533

Open
wants to merge 11 commits into
base: master
Choose a base branch
from
Open

Adding Gaussian Blur as 'hide_color' option #533

Changes from all commits
Commits
Show all changes
11 commits
Select commit Hold shift + click to select a range
4c917fa
Merge pull request #1 from marcotcr/master
palatos Oct 28, 2020
245d60c
Adding blur perturbation
palatos Oct 28, 2020
004df4e
no message
palatos Oct 28, 2020
c1c9218
Adding Blur
palatos Oct 28, 2020
67a4da6
Test
palatos Oct 28, 2020
9454010
Correction
palatos Oct 28, 2020
c5d604d
no message
palatos Oct 28, 2020
cde1879
no message
palatos Oct 28, 2020
e37ac97
Adding blur option to hide_color
palatos Oct 28, 2020
e8bcacc
Blur
palatos Oct 28, 2020
fdc13ba
Add skimage gaussian blur
palatos Oct 28, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 7 additions & 1 deletion lime/lime_image.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,6 +15,7 @@
from . import lime_base
from .wrappers.scikit_image import SegmentationAlgorithm

from skimage.filters import gaussian

class ImageExplanation(object):
def __init__(self, image, segments):
Expand Down Expand Up @@ -186,12 +187,17 @@ def explain_instance(self, image, classifier_fn, labels=(1,),
raise e

fudged_image = image.copy()
if hide_color is None:

if hide_color is 'blur':
fudged_image = gaussian(fudged_image, sigma=4, multichannel=True, preserve_range = True)

elif hide_color is None:
for x in np.unique(segments):
fudged_image[segments == x] = (
np.mean(image[segments == x][:, 0]),
np.mean(image[segments == x][:, 1]),
np.mean(image[segments == x][:, 2]))

else:
fudged_image[:] = hide_color

Expand Down