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

Modify the Dice loss #376

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

Modify the Dice loss #376

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
8 changes: 6 additions & 2 deletions training/loss_fns.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -9,6 +9,7 @@

import torch
import torch.distributed
import torch.linalg as LA
import torch.nn as nn
import torch.nn.functional as F

Expand All @@ -20,6 +21,9 @@
def dice_loss(inputs, targets, num_objects, loss_on_multimask=False):
"""
Compute the DICE loss, similar to generalized IOU for masks
Reference:
Dice Semimetric Losses: Optimizing the Dice Score with Soft Labels.
Wang, Z. et. al. MICCAI 2023.
Args:
inputs: A float tensor of arbitrary shape.
The predictions for each example.
Expand All @@ -38,11 +42,11 @@ def dice_loss(inputs, targets, num_objects, loss_on_multimask=False):
# flatten spatial dimension while keeping multimask channel dimension
inputs = inputs.flatten(2)
targets = targets.flatten(2)
numerator = 2 * (inputs * targets).sum(-1)
else:
inputs = inputs.flatten(1)
numerator = 2 * (inputs * targets).sum(1)
denominator = inputs.sum(-1) + targets.sum(-1)
difference = LA.vector_norm(inputs - targets, ord=1, dim=-1)
numerator = (denominator - difference) / 2
loss = 1 - (numerator + 1) / (denominator + 1)
if loss_on_multimask:
return loss / num_objects
Expand Down