Download experiments/losses.py from Angione-Lab/Shape2Force: direct link, hf CLI and curl.
- Browser
- Download file 1.33 kB
-
https://huggingface.co/Angione-Lab/Shape2Force/resolve/main/experiments/losses.py
- Command line
-
hf download hf://Angione-Lab/Shape2Force/experiments/losses.py
-
curl -L -o losses.py https://huggingface.co/Angione-Lab/Shape2Force/resolve/main/experiments/losses.py
1.33 kB
| """ | |
| Reconstruction losses. | |
| PerImageWFMRMELoss: relative magnitude error, each pixel weighted by its ground-truth magnitude divided by | |
| the mean ground-truth magnitude of the same image. | |
| WeightedMSEPlusDiceLoss: MSE with pixels above `threshold` weighted by 1 + pos_weight, plus a soft-Dice term. | |
| """ | |
| import torch | |
| import torch.nn as nn | |
| class PerImageWFMRMELoss(nn.Module): | |
| def __init__(self, eps=1e-6): | |
| super().__init__() | |
| self.eps = eps | |
| def forward(self, pred, target): | |
| t, p = target[:, 0], pred[:, 0] | |
| rel = (p - t).abs() / (t + self.eps) | |
| w = t / (t.mean(dim=(1, 2), keepdim=True) + self.eps) | |
| return (rel * w).mean() | |
| class WeightedMSEPlusDiceLoss(nn.Module): | |
| def __init__(self, pos_weight=3.0, threshold=0.2, dice_weight=0.5): | |
| super().__init__() | |
| self.pos_weight, self.threshold, self.dice_weight = pos_weight, threshold, dice_weight | |
| def forward(self, pred, target): | |
| w = 1.0 + self.pos_weight * (target > self.threshold).float() | |
| mse = (w * (pred - target) ** 2).mean() | |
| p, t = pred.flatten(1), target.flatten(1) | |
| dice = (2 * (p * t).sum(1) + 1e-6) / (p.sum(1) + t.sum(1) + 1e-6) | |
| return mse + self.dice_weight * (1 - dice.mean()) | |
| REC_LOSSES = {"rme": PerImageWFMRMELoss, "wmd": WeightedMSEPlusDiceLoss} | |