Multi-Task Learning with Additive U-Net for Image Denoising and Classification
This work addresses the challenge of stable and scalable multi-task learning in computer vision, though it appears incremental as it builds on existing U-Net architectures with a novel fusion method.
The paper tackled the problem of improving training stability and performance in multi-task learning for image denoising and classification by proposing an Additive U-Net that replaces concatenative skips with gated additive fusion, achieving competitive reconstruction performance and systematic task-aware redistribution of skip weights.
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.