Abstract:Accurate image segmentation remains challenging, particularly in generating sharp, confident boundaries. While modern architectures have advanced the field, many of them still rely on standard loss functions like Cross-Entropy and Dice, which often neglect the discriminative structure of learned features, leading to inaccurate boundaries. This work introduces Deep Discriminant Analysis (DDA), a differentiable, architecture-agnostic loss function that embeds classical discriminant principles for network training. DDA explicitly maximizes between-class variance while minimizing within-class one, promoting compact and separable feature distributions without increasing inference cost. Evaluations on the DIS5K benchmark demonstrate that DDA consistently improves segmentation accuracy, boundary sharpness, and model confidence across various architectures. Our results show that integrating discriminant analysis offers a simple, effective path for building more robust segmentation models.
| Comments: | Accepted to ICIP 2026 |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG) |
| Cite as: | arXiv:2605.14609 [cs.CV] |
| (or arXiv:2605.14609v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2605.14609 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Raül Pérez-Gonzalo [view email]
[v1]
Thu, 14 May 2026 09:25:36 UTC (11,556 KB)