← Search

Henrik Hult

2 accepted papers

2025

The Impact Label Noise and Choice of Threshold has on Cross-Entropy and Soft-Dice in Image Segmentation

CVPR 2025poster

In image segmentation and specifically in medical image segmentation, the soft-Dice loss is often chosen instead of the more traditional cross-entropy loss to improve performance with respect to the Dice metric.Experimental work supporting this claim exists, but how and why the two loss functions le…

Cited by 0SourcePDFScholar
2022

On Image Segmentation With Noisy Labels: Characterization and Volume Properties of the Optimal Solutions to Accuracy and Dice

NeurIPS 2022accept

We study two of the most popular performance metrics in medical image segmentation, Accuracy and Dice, when the target labels are noisy. For both metrics, several statements related to characterization and volume properties of the set of optimal segmentations are proved, and associated experiments a…