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

Update utils.py #6

Open
wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Update utils.py #6

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
37 changes: 37 additions & 0 deletions utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -173,6 +173,43 @@ def compute_mean_iou(flat_pred, flat_label,info=False):
mean_iou = np.mean(Intersect / Union)
return mean_iou

# This function is written by Ying @UIUC RSim group. In this case, we have assumed the batch size is 1.
def getCoordinates(predict):
pupilPixel = np.where(predict[0].cpu().numpy() == 3) # There are four values (0, 1, 2, 3) in the segmentation result arrays. Value 3 indicates the pixel for pupil.
# pupilPixel[0] hass the x-values of the pupil pixels, and pupilPixel[1] stores the y-values of the pupil
# pixel. They have an one-to-one correspondence. E.g., for pixel i, pupilPixel[0][i] is the x-value of this
# pixel, while pupilPixel[1][i] is the y-value of this pixel..

xMin = min(pupilPixel[0]) # This is the left-most pixel of the pupil.
xMax = max(pupilPixel[0]) # This is the right-most pixel of the pupil.
yMin = min(pupilPixel[1]) # This is the upper-most pixel of the pupil.
yMax = max(pupilPixel[1]) # This is the lower-most pixel of the pupil.
Comment on lines +183 to +186
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Can you do the indexing pupilPixel[0][i] instead of min or max this makes it easier to understand what you are doing.


delta = (xMax - xMin) - (yMax - yMin) # We want to use this variabl to decide the center of the pupil. If the difference is larger from the horizontal
# axis, we will use the x-value to decide the center of the pupil. Similarly, if the vertical difference is larger,
# we will use the y-value to decide the center of the pupil.

if delta >= 0:
xCenter = int((xMin + xMax) / 2)
xMin_yMin = min(pupilPixel[1][np.where(pupilPixel[0] == xMin)])
xMin_yMax = max(pupilPixel[1][np.where(pupilPixel[0] == xMin)])
xMin_yCenter = (xMin_yMin + xMin_yMax) / 2
xMax_yMin = min(pupilPixel[1][np.where(pupilPixel[0] == xMax)])
xMax_yMax = max(pupilPixel[1][np.where(pupilPixel[0] == xMax)])
xMax_yCenter = (xMax_yMin + xMax_yMax) / 2
yCenter = int((xMin_yCenter + xMax_yCenter) / 2)
elif delta < 0:
yCenter = int((yMin + yMax) / 2)
yMin_xMin = min(pupilPixel[0][np.where(pupilPixel[1] == yMin)])
yMin_xMax = max(pupilPixel[0][np.where(pupilPixel[1] == yMin)])
yMin_xCenter = (yMin_xMin + yMin_xMax) / 2
yMax_xMin = min(pupilPixel[0][np.where(pupilPixel[1] == yMax)])
yMax_xMax = max(pupilPixel[0][np.where(pupilPixel[1] == yMax)])
yMax_xCenter = (yMax_xMin + yMax_xMax) / 2
xCenter = int((yMin_xCenter + yMax_xCenter) / 2)

return xCenter, yCenter

def total_metric(nparams,miou):
S = nparams * 4.0 / (1024 * 1024)
total = min(1,1.0/S) + miou
Expand Down