Multi-Task Learning with Additive U-Net for Image Denoising and Classification
Fuente:
arXiv
Guardado en:
| Autores principales: | , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2026
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866915796501921792 |
|---|---|
| author | Lakkavalli, Vikram Sinha, Neelam |
| author_facet | Lakkavalli, Vikram Sinha, Neelam |
| contents | We investigate additive skip fusion in U-Net architectures for image denoising and denoising-centric multi-task learning (MTL). By replacing concatenative skips with gated additive fusion, the proposed Additive U-Net (AddUNet) constrains shortcut capacity while preserving fixed feature dimensionality across depth. This structural regularization induces controlled encoder-decoder information flow and stabilizes joint optimization. Across single-task denoising and joint denoising-classification settings, AddUNet achieves competitive reconstruction performance with improved training stability. In MTL, learned skip weights exhibit systematic task-aware redistribution: shallow skips favor reconstruction, while deeper features support discrimination. Notably, reconstruction remains robust even under limited classification capacity, indicating implicit task decoupling through additive fusion. These findings show that simple constraints on skip connections act as an effective architectural regularizer for stable and scalable multi-task learning without increasing model complexity. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_12649 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Multi-Task Learning with Additive U-Net for Image Denoising and Classification Lakkavalli, Vikram Sinha, Neelam Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning We investigate additive skip fusion in U-Net architectures for image denoising and denoising-centric multi-task learning (MTL). By replacing concatenative skips with gated additive fusion, the proposed Additive U-Net (AddUNet) constrains shortcut capacity while preserving fixed feature dimensionality across depth. This structural regularization induces controlled encoder-decoder information flow and stabilizes joint optimization. Across single-task denoising and joint denoising-classification settings, AddUNet achieves competitive reconstruction performance with improved training stability. In MTL, learned skip weights exhibit systematic task-aware redistribution: shallow skips favor reconstruction, while deeper features support discrimination. Notably, reconstruction remains robust even under limited classification capacity, indicating implicit task decoupling through additive fusion. These findings show that simple constraints on skip connections act as an effective architectural regularizer for stable and scalable multi-task learning without increasing model complexity. |
| title | Multi-Task Learning with Additive U-Net for Image Denoising and Classification |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2602.12649 |