Test-time Adaptation for Real Image Denoising via Meta-transfer Learning
This addresses the problem of adapting denoising models to real-world noise during testing, offering an incremental improvement over existing methods.
The paper tackles real image denoising by proposing a two-stage meta-transfer learning strategy for test-time adaptation, which outperforms other state-of-the-art methods on a real noisy dataset.
In recent years, a ton of research has been conducted on real image denoising tasks. However, the efforts are more focused on improving real image denoising through creating a better network architecture. We explore a different direction where we propose to improve real image denoising performance through a better learning strategy that can enable test-time adaptation on the multi-task network. The learning strategy is two stages where the first stage pre-train the network using meta-auxiliary learning to get better meta-initialization. Meanwhile, we use meta-learning for fine-tuning (meta-transfer learning) the network as the second stage of our training to enable test-time adaptation on real noisy images. To exploit a better learning strategy, we also propose a network architecture with self-supervised masked reconstruction loss. Experiments on a real noisy dataset show the contribution of the proposed method and show that the proposed method can outperform other SOTA methods.