Controlled Data Rebalancing in Multi-Task Learning for Real-World Image Super-Resolution
This addresses the challenge of handling different degradation patterns in real-world image super-resolution, though it appears incremental as it builds on existing multi-task learning paradigms.
The paper tackles the problem of task imbalance in multi-task learning for real-world image super-resolution by proposing a framework with coordinated advancements in task definition, imbalance quantification, and adaptive data rebalancing, achieving consistent superiority across all degradation tasks.
Real-world image super-resolution (Real-SR) is a challenging problem due to the complex degradation patterns in low-resolution images. Unlike approaches that assume a broadly encompassing degradation space, we focus specifically on achieving an optimal balance in how SR networks handle different degradation patterns within a fixed degradation space. We propose an improved paradigm that frames Real-SR as a data-heterogeneous multi-task learning problem, our work addresses task imbalance in the paradigm through coordinated advancements in task definition, imbalance quantification, and adaptive data rebalancing. Specifically, we introduce a novel task definition framework that segments the degradation space by setting parameter-specific boundaries for degradation operators, effectively reducing the task quantity while maintaining task discrimination. We then develop a focal loss based multi-task weighting mechanism that precisely quantifies task imbalance dynamics during model training. Furthermore, to prevent sporadic outlier samples from dominating the gradient optimization of the shared multi-task SR model, we strategically convert the quantified task imbalance into controlled data rebalancing through deliberate regulation of task-specific training volumes. Extensive quantitative and qualitative experiments demonstrate that our method achieves consistent superiority across all degradation tasks.